Evolution considered in the light of hybridization : lectures delivered at the university colleges of the New Zealand University, 1925

Lotsy, Johannes Paulus, Canterbury College, Christchurch, N.Z., 1925

This eBook is a reproduction produced by the National Library of New Zealand from source material that we believe has no known copyright. Additional physical and digital editions are available from the National Library of New Zealand.

EPUB ISBN: 978-0-908327-32-4

PDF ISBN: 978-0-908330-28-7

The original publication details are as follows:

Title: Evolution considered in the light of hybridization : lectures delivered at the university colleges of the New Zealand University, 1925

Author: Lotsy, Johannes Paulus

Published: Canterbury College, Christchurch, N.Z., 1925

Evolution Considered in the Light of Hybridization

BY J. P. LOTSY.

Lectures delivered at the University Colleges of the New Zealand University, 1925.

With Introduction and List of New Zealand Hybrids

by Da. L. COCKAYNE

Printed for CANTERBURY COLLEGE

by Andrews. Baty & Co. Ltd. Christchurch. N.Z,

1925

PREFACE

Early in the year 1924 Dr. L. Cockayne wrote to the Professorial Board of Canterbury College suggesting that Dr. J. P. Lotsy should be invited to lecture in New Zealand on his views of the evolution of animals and plants. The suggestion was warmly approved by the Professorial Board and the Board of Governors, and after the co-opera-tion of the other University Colleges had been obtained an invitation was sent to Dr. Lotsy. This was accepted and he left England on the 30th January, 1925, arriving in New Zealand on the 12th March. The three lectures contained in this small volume were written out during the voyage and were delivered for the first time before large audiences at Canterbury College on 17th, 19th and 25th March, 1926. They contained such an admirable summary from Dr. Lotsy's point of view of previous attempts to explain Evolution and contained so much new matter, that it seemed desirable they should be published with as little delay as possible. On being approached, Dr. Lotsy readily gave permission for them to be published by Canterbury College. The MS. of the lectures was copied and the printing put in hand at once. The lectures were afterwards delivered at the University of Otago, Victoria University College, and Auckland University College. They are now printed as they were delivered. Dr. Lotsy has himself read the proofs and has checked the references.

Dr. L. Cockayne has kindly written a short Introduction setting out the part played by hybridisation in the flora of New Zealand and has supplied the list of New Zealand plant hybrids which is printed in the appendix. The illustrations have been reproduced from photographs supplied by Dr. Lotsy.

CHAS. CHILTON,

Rector and Professor of Biology

Canterbury College, 6th June, 1925.

iii.

INTRODUCTION

by Dr. L. Cockayne.

As the idea of Dr. Lotsy being invited to deliver a course of lectures at the four New Zealand University Colleges came, in the first place, from me, Professor Chilton, who made all arrangements with regard to these lectures, is of opinion that it is of interest to tell how Dr. Lotsy’s visit to this country came about or, in other words, why the region would appear to him, or to anyone with similar aims, of special importance. To comply with this request of necessity requires a brief account of the part played by hybrids in the New Zealand flora, so far as the vascular plants are concerned.

It was in April 1921 that I had the good fortune while sauntering one evening in the beautiful forest near Elfin Bay, Lake Wakatipu, to accidentally find as the new growth, where certain trees had been burnt, a most diverse assemblage of sapling and seedling southern beeches (Nolhofagus), few of which were alike, the great majority matching no known species. This, at once, suggested hybridization, and later observations—assisted greatly by the State Forest Service —showed that southern-beech hybrids invariably occurred —frequently in abundance —throughout the range of the genus in New Zealand wherever certain speces were present. This case of hybridity led me into examining in the light of many years' experience the whole matter of wild hybrids in the New Zealand flora and I published a preliminary account of the subject in 1923 in The New Phytologhl. where a list of 130 supposed hybrids is given and a classification of these based on the degree of opportunity to cross.

While this investigation of hybrids was in progress, early on I wrote to Dr. Lotsy on the subject, as the one man most concerned therein, and he at once expressed his great interest and said how he hoped some day to visit the Dominion and see these hybrids of a more or less virgin vegetation in free nature. From this came the idea of his bringing before the biological students of New Zealand, and others interested, his views on evolution, by means of a series of advanced lectures, and I approached Professor Chilton on the matter and later the other Professors of Biology, the upshot of which is materialised in the following lectures.

At this point the question arises, why Is New Zealand a particularly good place, a region indeed almost unique, for Investigating matters of evolution P This question can be briefly answered by the facts, (I) that the floras and faunas have been far isolated from those of any other region for a very long period, (2) that a good deal of the vegetation is still primeval, or

V.

nearly so, (3) that the area, though comparatively small, yet embraces many climates, many land-forms, altitudinal belts from the coast to the perpetual snowflelds, and a diversity of plant-communities only to be seen elsewhere on an entire continent, hence (4) it follows that the whole is not too great for one person to grasp, and that in consequence the problems offered, though difficult enough, reach the minimum of simplicity, (5) that the flora consists of well-marked elements, the one purely New Zealand (amongst the seed-plants 80 per cent, are endemic) and the others with species or genera, identical or closely allied to those of other regions especially Australian, Malayan (including Polynesian) and subantarctic South American (Fuegian). Finally the recent action of man, in directly and indirectly altering the face of the land, has brought new plant-communities into being, some of indigenous and some of exotic species, and opportunities for hybridization have been afforded to species otherwise far-distant from one another.

To come now to the matter of wild hybrids, amongst the vascular plants at present 206 are known between Linnean species, but certainly a good many more will be discovered and within the Linnean species themselves there are undoubtedly a considerable number but these have been but little studied as yet. The 206 hybrids between species belong to 41 families and 85 genera. Hybrids indeed play a notable part both in the New Zealand flora and vegetation and the striking fact needs emphasising that, in many cases, it is not an occasional individual which is met with, but the hybrids occur in profusion. A list of such as are at present known by me is given in the appendix.

This knowledge of New Zealand hybrids, even insufficient as it must be, is a great contrast to that of a few years ago, when nearly all the two hundred and six were considered either “ good species ” or “ intermediates ” between such. In fact some of these hybrids amongst which there is no uniformity, e.g., Myrtus Ralphii, were considered not to vary at all. So evident, indeed to anyone, who accepts hybridization as the cause of the supposed “ variability ” of species, is the great “ diversity ” —to use Lotsy’s term —in many Linnean species, that in some cases it is at present impossible to find the real species, i.e. the true-breeding entities ! Thus, in my garden there are two collections of Hebe containing more than 120 forms, but each of these was collected from different plants selected from a mixed population occurring only on an acre or two, the parents of which for the most part I cannot even guess, while many are so distinct that any taxonomist, not knowing the facts of the case, would unhesitatingly publish them as “ valid ” species. Indeed not a few of the recognised species of Hebe are of this class. Taking other genera, too, who can separate Leptospermum scopariwn, manuka, into its true-breeding elements, or what is our real knowledge regarding the classification of Acaena, piripiri, or again how many of the so-called species of Celmisia are really such P In short, the whole flora demands a searching investigation in the light of hybridization. Here is work for many years to come for University students —work which cannot fail to throw light not merely on the composition of the flora but on methods of evolution, for the study of hybrids is intimately bound up with that of epharmony, of “ fluctuating variability ” and, in fact, of every kind of polymorphy. Not the least important of those subjects for research which await the trained, enthusiastic

vi.

investigator is detailed study of those curious species, one hundred or so in number, which have long-persisting juvenile forms, the origin of which is yet but mere speculation.

A preliminary in the investigation of hybrids, before actual crossing experiments can be made, is sowing the seeds of suspected hybrids. Thus Mr. W. A. Thomson of Dunedin from one head of Celmisia Traversii X verbascifolia procured seedlings in great diversity, and in my garden there is a mixed assemblage— no two alike—come from sowing seed of a supposed natural cross between Acaena inermis and A. Sanguisorbae, while twentysix years ago I showed that several forms of Hebe Traversii, each as much like the “ type ” as anything else, did not in the least breed true.

Diversity may well be expected amongst the progeny of hybrids, but constancy is what Dr. Lotsy’s theory demands, and this in wild nature is far more difficult to prove. In the case of a certain apparently constant form of Nothofagus which is frequently met with, a hybrid origin seems the most rational explanation. More interesting, and hardly a matter of speculation, is the case of Plagianthm cymosus which Dr. Lotsy and I recently studied in the lower Pelorus Valley, Marlborough. This so-called species is found only where a tidal river extends into a rich alluvial flat and in consequence Plagiantbus belulinus (lowland-ribbonwood), a rather small tree with dense crown, comes in contact with P. divaricatus, a twiggy divaricating-shrub, confined to a more or less saltish soil. The greater part of the hybrids are small trees with slender twiggy branches and leaves and inflorescences intermediate between the parents ; they are about as uniform as an ordinary invariable Linnean species. But there are occasional transitional forms between the above and the parents, so that the group as a whole resembles a “ variable species.”

Besides the question of natural hybrids. Dr. Lotsy had also in mind for study in New Zealand the behaviour of the many introduced plants and animals. In the case of the latter he expected to learn something concerning the segregation of definite races, or in what direction changes were tending or to what extent diversity had come about. As for the exotic plants it would be instructive to see if differences similar or different from those known in Europe had arisen. Or hybrids might even occur between the indigenous and the introduced as has happened in the case of the Australian Acaena ovina and the indigenous A. Sanguisorbae. Finally all matters connected with the breeding of domestic animals would be of interest, especially the case of the New Zealand Corriedale sheep.

In conclusion, enough has been said to show what splendid experiments, natural and artificial, are being carried out in this land of ours, and not only how they are a source of interest to our notable visitor, but how much material is to hand for our botanists and zoologists to investigate from many standpoints, ranging from field observations to cytology. In this allimportant matter of stimulating research and in turning our minds towards a field generally neglected but rich to excess. Dr. Lotsy’s visit should be epoch-making for New Zealand science.

L. COCKAYNE,

Ngaio, April 29th, 1925

vii.

LECTURE 1.

Different Attempts to Explain Evolution.

The idea that there has been a gradual development of the earth and of the living beings which people it is an ancient one. so that speculative evolution can be traced back to very remote times. Mod, in idea- of evolution however could not be developed before the idea of the species had been conceived.

This idea was most clearly expressed by Linnaeus : species tot numeramus quot creavit Infinitum Ens —there are as many species as were created by the Deity— i.e., the Lord created different kinds of individuals each of which, or a pair of them, was endowed with (he power to reproduce its kind faithfully. On the ground of this conception, it should be easy to find the species in the apparent chaos of living beings which people the world : all one had to do was to study the differences between various individuals and to unite all which possessed the same kind of characters to one species.

It was this principle which guided Linnaeus in all bis researches, but there is a meal difference between the conception (if a principle and its practical application. This Linnaeus soon appreciated. There are no identical individuals in nature, but similar ones only, and tliis fact alone suffices to make the species-conception an uncertain one. Linnaeus however, who was an excellent obarver, imagined that he was able to get around this difficulty. He saw that living beings are modified by their surroundings, but that these modifications need not In- inherited. For instance : grains of wheat collected on small plants, which grow under poor conditions, give, when sown in good soil, quite a respectable crop— '.<-.. the plants derived from them are much better developed than their parents. Consequently the fact that we do not tind identical, but only similar, individuals in nature could be so explained, that the circumstances under which different individuals grow up are never quite the same. In as much as such small differences are not inherited they of course do not affect the conception of the constancy of the species.

Such non-transmissible differences, which at the present time are designated modifications, were called varietales levissimae, unessential differences, and he forbade his students to study them : varietales levissimas non curat botanicus.

1

This standard was perfectly justifiable, as long as the only aim of the taxonomist was the finding of the immutable, created species. It became however soon apparent that the question was by no means such a simple one : that, on the contrary, there were transmissible differences within the groups of individuals, which Linnaeus considered to be good species. He himself gives an example of this. Within his species Brassica oleracea he points out that there exist a number of races, such as red cabbages, green cabbages, Brussels sprouts, etc., which reproduce their kind faithfully.

That there was a fundamental difference between non-transmissible and transmissible diversity within his species was clearly recognised by Linnaeus, as follows from the fact that, while he forbade the study of the former to his students, he recommends the study of the latter to them. Unfortunately he designated the transmissible differences within his species by a name, which resembled the one that he applied to the non-transmissible ones, far too much : he called them varietates, tout court, and recommended his students : varietates attente inspiciantur.

Varietates levissimae —non-transmissible differences—consequently should be neglected ; varietates —transmissible ones—should be carefully studied. No wonder that such similar terms for such fundamentally different things caused confusion, even in Linnaeus’s own mind, to such an extent that at a later date he wrote : Varietas est planta mutata a causa accidentali, ventis, calore, etc., where, of course, he should have written : varietas levissima. As a matter of fact, the difference between a varietas levissima and a varietas, between a non-transmissible and a transmissible difference, was soon forgotten and it became the fashion to unite all similar individuals to a species, without bothering about the question whether the differences between them were transmissible or not: a certain type was chosen as the ideal for each species, and deviations from this chosen type were either neglected or described as mere varieties of the species in question, while, subconsciously, one considered such varieties as non-transmissible, as temporary deviations from the species only, as individual differences, as Nageli liked to call them.

In this way the drawing of the limits between the species became a question of personal taste : one author considered a certain individual to be a mere variety of a species already described, while another one considered that very same individual as sufficiently distinct from it to describe it as the type of a new species. So it came about that there arose a belief in the so-called taxonomic insight, which allowed the happy possessor of it to distinguish at sight between a species and a variety. Elects possessed this insight, which ordinary people lacked ; unfortunately the latter thought themselves to be so distinguished, so that there arose a constant dispute about the question which species were good species and which were not, a dispute which, I fear, is even now not yet entirely extinct.

The question whether constant species do exist or not, can of course be decided in one way only, by putting the inheritance or noninheritance of such characters as are considered to be essential, to the test, a test with which Linnaeus himself already made a modest beginning in the case of his races of cabbages.

2

That one did not apply this test universally, had a very good reason. Of the many kinds of living beings which people the earth but a comparatively small number is suitable material for experiments—mice are quite good in this respect, lions and tigers considerably less so, and whales may be called quite unsuitable.

Plants however are usually excellent material, so that it is rather curious that not a botanist, but a dealer in silk goods in Lyons, Alexis Jordan, put the question of the constancy of the species to the test. About the middle of the 19th century — i.e., a very long time after Linnaeus, almost 50 years even after Lamarck’s attack on the constancy of the species—Alexis Jordan put the question at issue for the first time in the right way : “ I see,” he said, “ within Linnaeus’s species appreciable differences : what is the nature of these differences, are they transmissible or are they not ? ” And he took the only measure suitable to give an answer to this question ; he sowed the seeds of self-fertilized individuals and studied the progeny of them so obtained. He soon found that there really were transmissible differences within Linnaeus’s species, and as transmissibility of the characters is, according to Linnaeus himself, the criterion of a species, he concluded, quite logically, that there existed within Linnaeus’s species different types which, in as much as they transmitted their peculiarities faithfully to their offspring, had to be considered as species, so that in regard to them the linnean species became a genus.

These real species he called petites especes or microspecies and, after a study of thirty years, he could show that within a single species of Linnaeus, within his Draha verna, there existed no less than 200 of such microspecies. So it seemed as if the experiment’corroborated Linnaeus’s hypothesis of the existence of constant species, although the recognition of them was much more difficult than Linnaeus had imagined it to be, and Jordan was therefore perfectly convinced that he had discovered the real created species.

One point, an important one, however, was not fully appreciated by Jordan. His experiments had taught him that while certain individuals transmitted the whole complex of their characters to their offspring, others distributed this complex over their progeny in such a way that each member of the latter obtained but a larger or smaller part of it, or, as we now express it, that such individuals segregate. He had also already recognised that these segregating individuals were hybrids. This discovery practically—although Jordan did not recognise this—put an end to the species-conception, because it is the criterion of the species, that any member of it has had no other ancestors than such as were like itself, while Jordan had, by his experiments, proved that individuals exist whose progeny differs from them, individuals therefore which can give birth to individuals with other characters than they themselves possess.

That Jordan, excellent observer and excellent critic as he was, neglected this important fact, that he did not recognise the significance of the existence of such fertile hybrids for the species question, was due to his firm belief in the strict truth of every word of Scripture, as results from the fact that after having expressed his satisfaction of having found what he supposed to be the real created species, he says : “ Religious tradition, although no part of science, has yet to serve

3

the human mind as a compass to guide it in its researches, and of the point in question revealed truth speaks an unmistakeable language, for vve read in Genesis : “ And God said, Let the earth bring forth grass, the herb yielding seed after his kind, and the tree yielding fruit, whose seed was in itself, after his kind.”

All the same, he energetically refutes the opinion of Decaisne that the latter had proved by his sowings of cultivated pears, which gave a heterogeneous progeny, that plants may vary, and concludes quite rightly from that result, that the pears of which Uecaisne sowed the seeds were hybrids.

But that thu»i* individuals, which gave rise to a progeny differing from them, annihilated the conclusion which he drew from his experiments, that his microspecies had had from the very beginning ancestors exactly like themselves, that the ej : fertile hybrids attacked the species-concept at its ' on account of his very belief in the strict truth oi every word ol Genesis. We will not judge him for that, when we remember how. as late as the beginning of last century, Scottish peasants killed a mule because hybrids were contrary to tile intentions of the Creator, or stranger still, that the celebrated Botanist Carl von Nageli declared in 1838 the numerous hybrids within the genus Cirsium as products contrary to nature, or at least as unnatural products, while even in our days there still sometimes exists in scientific circles a certain prejudice against hybrids, which is the more regrettable as every one of us —as the diversity of our children clearly shows —is a hybrid himself.

In regard to the value to be attributed to hybrids, Linnaeus was ahead of Jordan, because in his later days he no longer dogmatically upheld the doctrine of the constancy of species, as results from the fact that he changed his original saying : Species tot numeramus quot ah initio creavit Infinitum Ens, into in principio creavit Infinitum Ens, thus giving evolution a chance, a chance moreover by means of hybridisation. Unfortunately, in working out this idea, Linnaeus imagined quite curious, even impossible, hybrid combinations as, e.g., the wellknown one between a Saponaria and a Genliana, and this proceeding has doubtless contributed to the little esteem in which hybrids were held and prevented their recognition as a principle of evolution. All the same, one of Linnaeus’s hybrids, Tragopogon pratensis X porrijolius, really is a hybrid, which I myself have been able to re-obtain.

The first period of the investigation of living organisms, the period of the exclusive study of their differences, seemed therefore, by Jordan’s experiments, to corroborate the original idea of Linnaeus, viz. : the existence of numerous, created, constant and therefore immutable species, and this belief became a dogma of science.

The course of scientific progress unfortunately is not a straight one ; it has many tributaries, a number of which are apt to run in a backward direction.

Before Jordan, two men, whose names are now household words, had also investigated the question at issue. One of these, Lamarck, was both a taxonomist and a philosopher, who in 1809 already attacked

4

the dogma of the constancy of the species. His work was known to Jordan who, clearly recognising its hypothetical foundation, not only criticised but even ridiculed it.

The other, Mendel, as devout a Catholic as Jordan himself, penetrated deeper into the question of hybridisation than the latter and concluded from his experiments that Lamarck’s views could not be right, so that, if his work had not remained unknown to Jordan, the latter would probably have lent a sympathetic ear to it and obtained a better insight into the importance of hybridisation than he actually did.

Lamarck was in 1809, when he published his Philosophic Zoologique, the book which contained the first attack on the dogma of the constancy of the species, already past middle age. He was a systematist of great repute, whose works embraced the whole living kingdom—plants as w'ell as animals—and who had gained great fame by his careful work in the then very little known group of invertebrate animals. His system was, in many ways, far superior to that of Linnaeus, No man could be better qualified than he to consider the species-question.

He opens the question by pointing out that even if species were related to one another, w r e must remain in ignorance of these relations if we continue to study their differences only ; the only possible way to discover such eventual relations is to study the points which species have in common.

This study soon taught him that species are not sharply delimited, that on the contrary there are intermediate forms between them, and these intermediate forms he considers as transitional ones. So long only as scanty material from a region, limited in extent, is at our disposal, we imagine that two species are very distinct and well defined, but the more individuals we gather from different regions the more we recognise that the limits between them are hazy, the more also we recognise that our subdivision of Nature into species, genera, families, orders and classes is an artificial one, that it is we ourselves who create these groups and that in reality there are only individuals and nothing nut individuals in Nature.

According to Lamarck's views, species do not exist at all; what does exist is a single created kind, or a few created kinds of living matter, which presents itself to us at different periods under a different guise—now as a reptile, then as a bird, subsequently as a mammal, and finally as a man, although, in order not to hurt religious feelings, he takes care to add in regard to man, that one could imagine this, if it were not contrary to the teachings of the Church.

Such a new, he himself says, may at first sight appear to be untenable, as we are wont to consider birds and mammals, for instance, as very distinct and fundamentally different groups of living beings, but if we remember, he says, that two kinds of animals exist both of which lay eggs and yet suckle their young, Echidna and Monolrema, the latter of which even possesses the bill of a bird, we recognise that there exist even between two groups apparently so distinct, as birds and mammalia, transitional forms.

5

What we consider to be species are, according to Lamarck, by no means separate units in the sense of Linnaeus and Jordan, but merely temporary aspects of the life-substance which all living beings have in common.

When Lamarck, notwithstanding this view, often speaks of species and their variability, he does so more as a concession to established custom than from conviction.

On the contrary, he is so firmly convinced of the unity of Life, that he decidedly rejects the opinion that species can become extinct, with the only exception of such forms as are hunted by man. Extinction, to his way of thinking, is but simulated ; the form, which has to all appearances been annihilated, has in reality only been transformed to one of the next period in the earth’s history— Lamarck thoroughly believed in the Greek doctrine iravTa..pet, everything changes but nothing gets lost, the substance of life as little as anything else. It is therefore, strictly speaking, not right to consider Lamarck to be guilty of the doctrine of the inheritance of acquired characters, at least not when we ascribe to inheritance any considerable degree of permanency ; according to Lamarck, nothing is permanent.

How then does Lamarck, who distinctly denies the existence of separate specific units, imagine that evolution, for he is the first one in modern times who offered a hypothesis of evolution, took place.

He shows, and this was a great progress at the time when he wrote, that all living substance is irritable, and he imagined—that it was merely an imagination we shall presently see—that the stimuli of the surroundings were able to influence the life-substance in such a way, that it responded by forming the organs and other properties which were essential for its existence in the near future. It is therefore the surroundings which, by the response they obtain from the lifesubstance, cause the form and the organisation which the organism needs at a certain time ; it is the surroundings themselves which adapt the organism directly to its surroundings or, as Lamarck likes to express JL, it is the needs of the organism which creates the required organisation and form-

To quote an example : the giraffe obtained its long neck by its continued efforts to reach the leaves on shrubs in the dry regions in which it has to live, the duck obtained its webbed feet by the continuous efforts of its ancestors to tread the water, land animals their lungs by their desire to gain the land ; in one word, one can not be too careful in the choice of one’s parents. Although such views to most of us appear to be phantastic, one can, as Lamarck did, offer some facts in their support.

It was of course known, and Lamarck brings this forward, that the form of a number of plants is very different under different circumstances ; it is nett easy to recognise in the water form and the land form, and in the still more peculiar dune form of Polygonum amphibium, merely different aspects of the same kind of organism, of one and the same individual even, as Massart showed. Yet, it is this very Polygonum amphibium which shows, in the clearest possible manner, Lamarck’s mistake in considering such modifications to be transmissible

6

because one can take an individual of the land form and merely by submerging it in water force it to take the shape of the water form, and then plant it out in the dunes and see this very same individual take the typical shape of the form which grows in the dunes, and not only its shape but its peculiar organisation also.

But not always is the change so easily brought about; many plants of the Alps transplanted into the plains retain the shape which they acquired at their elevated habitat for a considerable time, but —and this is conclusive—we have only to sow their seeds to see that their progeny retains none of the characters which their parents acquired during their long stay in the Alps.

Nor is the idea, expressed by Lamarck, that animals with lungs can arise from such as used to breathe by means of gills, so phantastic as it might appear to be at first sight.

As Lamarck points out, we all know that the frog, which lives on the land and breathes by means of lungs, passed through the tadpole stage during which it lived in the water and breathed by means of gills. Lamarck however overlooks the decisive point that the tadpole passes into the frog-stage without having felt the stimulus of the life on the land.

As a matter of fact, we have not a single irrefutable proof, in my opinion, for the inheritance of stimuli of the surroundings, for a formation of permanent engrams in the life-substance, as Semon, later, so well expressed it.

All the same, a considerable number of very able naturalists are openly or in the secret recesses of their hearts, adherents of the Lamarckian doctrine.

This attitude, of which I shall speak more in detail in one of my next lectures, is, it seems to me, mainly due to an incorrect view as to what constitutes heredity and to an exaggerated idea as to the influence of time in the acquisition of transmissibility. I will illustrate this by two examples, which were actually offered as proof against my views or at least as points which spoke against it.

The first concerned a woman, who by a faulty construction of her foot had acquired a callus at an unusual place. A child born to her had a callus at a corresponding place, and this was considered as a case of inheritance of an acquired character. Now this is quite a mistaken idea : it is not the callus which is inherited but the power to produce one, and the whole of our skin possesses that power, so that it is merely a coincidence that the child possessed a callus at a spot corresponding to the one where the mother had one. Besides, the whole case is very suspicious and may very well have been one of mistaken identity—e.g., the abnormality on the child’s skin may not have been a callus at all, because even if such an abnormality were inherited by engramformation it would be very improbable that this particular child would show it, if we keep in mind how many different combinations are possible among the children of a couple, one of which only shows the abnormality.

To illustrate the mistaken conception of the saying that a character is inherited instead of the power to produce it, the following imaginary case may be of service. Suppose the deftness to pick pockets were

7

transmissible, we might say that a child born from parents with that deftness would be able to pick pockets more easily than a child born of less gifted parents. But nobody would think of assuming that such a child would be born with a picked purse in its hand, which indeed would be the analogon to a child appearing in the world with a callus, before any stimulus could have caused its power to produce one to come into action.

The other point, the exaggerated importance attached to a considerable amount of time being able to bring about the inheritance of acquired characters, may also be illustrated by one or two examples. To my way of thinking, it is just the length of time which Lamarck assumes to be necessary to bring about this transmissibility which forms a serious objection to his views.

A drowning person certainly feels the need of the ability to swim, but is drowned because he lacks this ability. And even if one would assume that the need felt gave the first impulse to the formation of an engram in his germplasma, he is unable to transmit this impulse for further improvement to his progeny, because his near death cuts him off from the possibility of begetting any progeny at all. 1 know very well that this is an exaggerated example, but I think that it does illustrate the principle ; it is just the long time required to produce the useful property which is so serious an objection to belief in the applicability of Lamarck’s hypothesis.

One frequently forgets that according to Lamarck's conception only suoh characters are formed by the influence of the surroundings as are useful to the organism. This would allow us to assume that man—supposing that Adam and Eve were white people—had acquired in Africa the black skin of the negro, through the influence of the tropical sun and as a protection against it. When, in a recent discussion. 1 expressed considerable doubt of white men being able to acquire a black skin under any circumstances. 1 was answered that what might not be possible in a short period might happen in a period extending over some millions of years.

This conception I consider erroneous ; the black skin is either a necessity to man living in Africa or it is not; if it is, acquisition after some millions of years is an impossibility because those who had to acquire it will have died long before they could get so remote a progeny, and if it is not, they will never acquire it according to Lamarck’s doctrine.

There is another point of importance which speaks against Lamarck’s views, to wit, the coexistence of very simple and very highly organised organisms. If an amoeba-like Urplasma could be transformed gradually by the stimuli of the surroundings unto man, by way of such subsequent stages as reptilia, amphibia, birds and numerous kinds of mammalia, w'hy then have not all amoeba-like urplasmata been thus transformed—how is it that part of them escaped the influence of those stimuli so completely that they did not advance any further than present-day amoebae have ?

Lamarck himself felt this difficulty so strongly that he did not hesitate, in order to get around it, to assume that generatio spontanea continued up to the present day, for which assumption there of course is not a particle of proof.

8

It further deserves notice that Lamarck also had to call in the aid of hybridisation to explain certain facts.

Talking about the origin of our different breeds of domestic dogs he says that these doubtless arose from a single wild race, the individuals of which were all alike and closely related to the wolf, and then he continues

" That race, in which there were no differences between the individuals pertaining to it was gradually distributed by man over the world, and so it came under the influence of different climates and of different circumstances which caused remarkable changes in it, thus creating different peculiar races. Moreover,as man for the sake of commerce or for other reasons travels over great distances, he brought into certain densely populated districts, such as large cities, a number of these races from very different countries, and from the crossing of these, there arose successively all those races of dogs which we now possess."

Is it, if one has to assume crossing as the cause of the origin of our present-day breeds of dogs, not more logical, and certainly more according to facts, that our dogs did not arise from a single wild species of Canis, but from the crossing of more than one ?

Lamarck did not succeed in getting his theory accepted, and this is quite natural because his hypothesis was mere speculation, without any support by facts.

Every organism doubtless is the result of its constitution and of the stimuli of its surroundings, but—happily for all of us—it is the constitution only which is inherited ; that is best shown by such a clear rase as that of Polygonum amphibium, in which the individual responds immediately to changes in the surroundings in a readily observable manner. Unfortunately this is not always the case, and this makes it so difficult to decide which characters are transmissible, because they are constitutional, and which are not. because they are mere modifications due to nurture.

While Lamarck, by his absolute denial of the existence of species which he considered merely as temporary modifications of a common life substance, was the direct antipode of Linnaeus, Darwin, who fifty years after Lamarck opened a new attack on the dogma of the constancy of the species, took up a less extreme position. It is rather staggering that Darwin, who called his book “ The Origin of Species,” did not define what he meant by a species, but merely says that every naturalist vaguely knows what should be understood by this term. There is no doubt that the Swiss botanist Moritzi expressed the standpoint which biologists took, and not infrequently still take, much more correctly, when he said : “We have all accepted the species, without knowing what we really mean by it.” Notwithstanding his expressed aversion to define his species, Darwin continuously uses this term, says that species are variable, in a transmissible sense, and that it is safest to assume that as a rule all deviations from type are transmissible, that non-transmissibility is the exception. Varieties he considers to be initia of species : “ Varieties,” he says, " are incipient species.” Because he considers each deviation as usually transmissible, he con-

9

tinuously oscillates between the opinion that larger deviations so-called sports, and that smaller ones, so-called modifications, are the material with which evolution operates.

Darwin apparently does not see that a variable species—variable in the sense of transmissible variability—is a contradictio in terminis, that it is supposed constancy, and constancy only , which was the cause of the assumption that there are species, and this faulty way of expressing oneself has been retained up to the present time, one still speaks of the variability of species.

This, I take it, is unjustifiable. There are in Nature perpetually constant types or there are not; if the former exist, species exist; if they do not, there are no species in existence either ; one is not allowed to designate different conceptions by the same name.

What Darwin really meant to say is that individuals are variable in a transmissible sense, but as he transferred this conception to those groups of individuals which he considered to be species, he mistook diversity within those groups, within his supposed species, for variability of those species.

I do not believe that even one of the many examples which Darwin gives of variability can stand the test of present-day methods, or would be accepted by any present-day naturalist after careful consideration.

No wonder, therefore, that one has disputed during decennia about the question what Darwin really understood by variability, and to which kind of it he attributed the greatest influence in Evolution. There exists however a letter of Darwin which makes it perfectly clear, that in the only case in which he intended to test variability by way of experiments, he mistook segregation of hybrids for variability.

In a letter addressed from Down, February, 1876, to Dr. Gilbert the well-known agricultural chemist, he writes :

" It is admitted by all naturalists that no problem is so perplexing as what causes almost every cultivated plant to vary, and no experiments as yet tried have thrown any light on the subject. Now for the last ten years I have been experimenting in crossing and self-fertilising plants, and one indirect result has surprised me much—namely, that by taking pains to cultivate plants under glass during several successive generations under nearly similar conditions, and by self-fertilising them in each generation, the colour of the flower often changes, and what is very remarkable, they become in some of the most variable species, such as Mimulus, Carnation, etc., quite constant like those of a wild species."

It is now of course perfectly clear that the plants with which Darwin began his experiments were hybrids, and that he mistook the segregation of those hybrids—which before him had already been observed by Jordan and Mendel, with whose works he. unfortunately, was not familiar—as evidence of their variability. That he did so. indeed, results clearly from the sequence of his letter to Dr. Gilbert, e.g. :

" This fact and several others have led me to the suspicion that the cause of variation must be in different substances absorbed from the soil by these plants when their powers of absorption are not interfered

10

with by other plants with which they grow mingled in a state of nature. Therefore my son and I wish to grow plants in soil entirely, or nearly entirely, destitute of all matters which plants absorb, and then to give during successive generations to several plants of the same species as different solutions as may be compatible with their life and health. And now can you advise me how to make soil approximately free of all the substances which plants naturally absorb 1 ”

Apparently this experiment has not been carried out, and it is certainly regrettable that Darwin, who in 1876 came so near to discovering the cause of so-called variability, had no knowledge of Mendel’s experiments which had been published thirteen years earlier. If he had known Mendel’s results, the cause of the result of his own repeated self-fertilisations would at once have become clear to him, and this might have completely changed his views as to the cause of evolution. Perhaps not, however, for Darwin knew that the Russian zoologist Pallas had already assumed that crossing was the cause of the supposed variability of our domestic animals, and therefore of the formation of our various breeds of them, although he did not extend this idea to the different forms of wild animals, because he, also, considered hybrids as unnatural products.

Darwin however opposes Pallas's ideas, as he is convinced that many domestic animals—as an example he mentions, among others, the various breeds of poultry—have descended from a single wild species. Even if this were so, it would by no means exclude the action of crossing, as Linnean species are not groups of identical individuals, but of genotypically different ones, so that crossing within the limits of such a species may have considerable effects. Moreover, lamin a position to affirm that Darwin was certainly mistaken in his assumption that our domestic breeds of poultry have been derived from a single wild species. Experiments began by Mr. Houwink, and continued by me, have shown that at least three wild species have taken part in the formation of our domestic poultry : to wit, Gallus bankiva occurring all over south-eastern Asia, Gallus sonnerati of British India, and Gallus varius of the Malay Archipelago. From the first of the two latter species our poultry obtained the silver factor which occurs in so many breeds which we call silver ones, from the latter the dominant black factor which characterises a number of other breeds.

We have seen already that Darwin did not recognise crossing as the real cause of the cases which he considered as variability. I may now add that even shortly before his death he expressed himself that the cause of variability remained obscure and must be considered to be a very remote and indirect one. It is however for a valuation of Darwin’s Theory of but secondary importance that he has not found the cause of the diversity which he observed. Of course nobody can observe anything but diversity, whether the observed diversity is due to variability, to crossing, or to any other cause can only be with certainty decided by experiment.

The chief point was that Darwin saw that individuals, which usually were united to one single species, could differ from one another and that this diversity was partly, at least, independent of their surroundings, or, as he expressed it, that they varied in all possible

11

directions, in such as were indifferent in respect to their surroundings, in such as were favourable in this respect and in others which were unfavourable.

As Darwin had recognised already for a considerable time that breeders made use of this diversity to adapt domestic products, by severe selection of those which best suited their purposes, to their requirements, he conceived the idea that, if an analogon to the breeder’s selection could be found in nature, this might possibly be the principle which adapted wild animals and plants to their own needs.

Darwin’s theory is therefore not primarily, nor even chiefly, a theory of the origin of species, but really a theory of the origin of adaptations. With Lamarck, he was convinced that only those organisms were able to persist—and consequently to make on us the impression of being more or less stable species—which possessed the power to adapt themselves to changes in their surroundings, and he also agreed with Lamarck that it was the changes in these surroundings which caused the changes in the organisms, the variability therefore which led to the formation of different species.

Not directly, however, as Lamarck thought. The power to react, always in a manner favourable to the organism, to stimuli emanating from the surroundings seemed to Darwin to be of too mystical a kind. How then could adaptation come about in an indirect way ?

We all know that Darwin and Wallace, independently of one another, came to what they considered to be the right solution, after reading Malthus’s Book on Population.

As many more individuals are born than remain alive, he assumed—and this at first sight appears to be perfectly legitimate—that those which were best adapted to their surroundings—those which fitted these surroundings best—were most likely to survive, and would gradually suppress the less fitted ones. In this way there could arise in the chaos, caused by the all-sided variability, a directing force pushing them in the direction of adaptation.

Seductive as this view may appear, two circumstances should not be forgotten : firstly, that the maximum of destruction takes place at so early an age, mostly even during the stage of eggs or seeds, that none of the characters between which selection might take place has yet been developed ; and, secondly, that such lower organisms as multiply by fission only, and such higher ones, as many apogamous ones, the progeny of which does not vary, yet are perfectly able to survive.

Doubtless a large amount of destruction takes place in nature, but this destruction by no means is limited to the less fitted ones ; we need only think in this respect of the millions of small organisms which are swallowed at one gulp by a whale. While travelling a couple of years ago in the Yellowstone Park in America, where the woods of the lodgepole pine are extremely dense, I was much struck by the difference in appearance between the trees near the roads and those further in the forest, many of which succumbed by the shade thrown by their companions. Doubtless a much larger percentage of the trees along the roads survived than of those inside the forest, not because the former had a better constitution, but because they occupied a better position, just as in a railway accident the feeble may survive and the strong be killed.

23

Many an apparently beautiful adaptation has been shown to be non-existent, such as, for instance, the adaptation to insect-pollination of the apogamous dandelion.

Doubtless we have greatly exaggerated the effectiveness of many so-called adaptations. By ingenious devices, stretching over a number of years, regarding the effectiveness of imitation of supposedly unpalatable butterflies by supposedly palatable ones, Tower could show that in Mexico the imitators and the examples were eaten just as frequently as the supposedly palatable ones. Nor were various other kinds of imitation —for instance, the well-known one of a dry leaf when the butterfly is at rest —any more effective, so that he, doubtless rightly, concludes that it is not fitness but a lucky position—just as in the case of the pine trees at the border of the roads —which gives the best chance of survival.

Notwithstanding the fact that Darwin’s theory was as little built on a solid foundation as that of Lamarck, notwithstanding the fact that both had to put the causes of their supposed variability in the remote and uncontrollable past, Darwin succeeded where Lamarck failed, so that after him the belief in Evolution has become universal. It is true that there are still Lamarckians, that even a neo-Lamarckism has developed, the great majority of naturalists however were of a Darwinian mind and only differed from one another as to the ways and means by which all-sided variability caused Evolution, especially whether large variations, so-called sports, or small individual variations, were most effective, whether Evolution took place per saltum or more gradually.

It was de Yries who tried to decide this point by experiments. Darwin already had, by following—unknown to him —Hippocrates' lead, advanced the hypothesis that the reproductive cells of organisms were nothing but an aggregate of minute particles—of pangenes, as he called them—which had been sent by the different kinds of cells which compose the organism to their reproductive organs. Each pangene consequently was practically a germ, a seed of a particular kind of cell; the reproductive cell was supposed to contain a complete set of all the germs of all the different kinds of cells which composed the organism.

Some such kind of hypothesis Darwin needed to explain the transmissibility of the effects of use and disuse in strengthening or weakening organs, which he, following Lamarck in this respect, assumed to exist. Darwin assumed these pangenes to be transported chiefly by means of the blood. Gallon tested this point experimentally by transfusion of blood of coloured rabbits into white ones, and as he found that this had no effect on the colour of the Utters born after transfusion, the result did not support Darwin’s views ; although of course one is not allowed to say that this experiment was so decisive as to disprove his views. De Vries improved Darwin’s hypothesis of pangenetio inheritance considerably by the assumption that not only each kind of cell but each property of the organism was due to the action of a particular pangene, which he imagined to be a minute, invisible unit of Ufe. In this way the pangenes became the bearers of the trans-

13

missible properties of the organisms, which, by a further assumption, had no need of such a way of transportation as Darwin had assumed lo be necessary,

This assumption was that the nucleus oi the zygote, from which the organism aro lj contained tin- whole set of pangenes present in that organism and notwithstanding its numerous divisions, transmitted it to all daughter nuclei, because each division of the nucleus was supposed i of all the pangenes it contained.

The cytoplasm, according to de Vries's conception, also ensistot pangenes, hut does not necessarily contain a complete set of these. because it is supposed to he Formed by hut pail of the pangenes of the as, u he hj left the nucleus, after having divided, bo as not to impair the compli 'in set. The kind of pangenes present in the cytoplasm is different in the case of each kind of cells present in the ii-iii. ..ml it i> this difference in pangene composition which is at the la,limn of the differentiation into different kinds of cells, into different kind- of organs subsequently, of each higher developed living being.

Attractive as this speculation is in many respects, it assumes rat lea- much, to mention one point only—it assumes besides the division of all pangenes during ordinary karyokinesis a division of part of the pangenes only whenever cytoplasm of a certain kind has to he formed, while no cytological phenomenon is known on which the latter assumption might he hased.

If we therefore abstract this part from de Vries's assumption, the rest, that the nucleus contains certain definite pangenes. is a very attractive one. because, as de Vries points out. it opens the way to explain the great diversity in nature by the assumption of numerous different combinations of a comparatively small number of pangenes, ceteris paribus, comparable to the large number of very different hooks which can he printed by means of the 2b letters of our alphabet.

This idea might, you may think, easily have led to the further assumption, that the cause of these numerous recombinations was hybridisation, but such an assumption would not have been sufficient, because de Vries supposes each pangene to be organised — i.e., to be the smallest living particle imaginable, so that if nothing but crossing ,1 the bottom of evolution, one would have to assume that all present-day pangenes hail been present already in the urplasmata, an assumption which would be highly improbable.

De Vries was therefor forced to imagine some kind of new formation of pangenes. and so he did. He assumed that progressive species-formation the only one which, to Ins way of thinking, causes evOltrtion was hased each time on the addition of a single new pangene t" ffie set lire at in the organism, this addition coming about by an action oi janism itself, by that organism forming that new pangene, the effect of which action he called mutation.

As lliis term has later on been used in a slip-shod manner, frequently [or inert', even for only supposed, alterations, without any reference to then cause so thai even the efiecl of crossing has been called mutation, it is absolutely i o point out thai the only legitimate use of the word, when we tali of the origin of so-called new species is to apply it to new formation of pangenes.

25

According to de Vries's views each species, each micro-species, for it is these which he calls species, differs from its mother-specic.-hy one single pangene, which was added, spontaneously, to the set of pangenes of the mother species by its own action This, of course, was a mere hypothesis, ready made, before de Vries began to look for experimental proof of it.

Almost a hundred species, he himself says, were tried before he found one which fulfilled his expectation.-. This of course is already a weak point in his theory: the most important question however i.s whether, as de Vries thinks, the Oenothera Lamarckiana experiments really did prove the existence of mutation

He really showed that self-fertilised Oenothera Lamarckiana gives birth to new forms, some of which are as constant as 0. La itself : forms 1 agree with him hen—which have as much, or as little, right as 0. Lamarckiana to the name of species. These forms he claimed furnished experimental proof of his mutation hypothesis. We all knowhow rapidly de Vries's mutation theory was accepted, which was due to the fact that all of us behoved in the existence of transmissible variability, but that none ol us had evei been abTe"t7)"prove"lte existence, so that we felt the greatest admiration for him who had at last succeeded.

It is a general rule that critical reading is not favoured by sympathy with the results obtained, and so it could happen that it was generally overlooked that de Vries's experiments in reality brought no proof whatever for the real existence of mutation.

The Lamarckianas with which he began his experiments had not been constant, and subsequently been induced by the way he treated them, to mutate, but they threw from the very beginning the aberrant forms, which de Vries merely fiilh>l mutants, and each and every Lamarckiana plant still throws them.

This forced de Vries to an additional hypothesis—to wit. that 0. Lamarckiana bad formed new pangenes, which manifested themselves as mutants already when he began his experiments—which additional hypothesis of course already annihilated his claim that he had proved that 0. Lamarckiana mutated.

Somewhat later, after it was shown that 0. biennis and other so-called species of the .subgenus Onagra of Oenothera threw similar aberrant forms as 0. Lamarckiana did, he was even forced to assume that the new pangenes which manifest themselves as mutants at the present time had been formed already in that remote time, when the initial forms of the sub-genus Onagra arose.

By the assumption of such remote premutation periods, the cause of the visible variability was again —mutatis mutandis —as by Lamarck and Darwin, put into the uncontrollable past. Moreover, as no cause could be given for the formation of new pangenes, de Vries’s hypothesis was in reality a hypothesis of generatio spontanea of pangenes, de Vries’s species-forming mutation is, just as Jordan’s hypothesis of the origin of species, a hypothesis of creation : Jordan considers the species, de Vries the pangenes, as created—i.e., spontaneously arisen.

15

That de Vries should have proved the existence of mutation is an opinion not in accordance with facts. It is finally of importance to point out that de Vries says, “ New species arise as hybrids,” because so few of the gametes formed by 0. Lamarckiana are supposed to contain the new pangenes that it is almost certain that they will be fertilised by gametes not containing these.

When we now look back on the theories of Lamarck, Darwin and de Vries, we see that they have two points in common :

1. The cause of the supposed variability is put into the remote past.

2. Hybridisation is used as an additional hypothesis, while the only cases in which the supposed variants were put to the test — e.g., by Darwin, in the case of his Mimuli and Carnations, and by de Vries in the case of his Oenotheras—it was subsequently proved that the plants supposed to be varying, were in reality heterozygotes or, in other words, hybrids, which segregate in a mendelian or other way.

16

LECTURE II.

Modern Attempts to Prove the Existence of Transmissible Variability.

We have seen in the former lecture that neither Lamarck, nor Darwin, nor de Vries has proved the real existence of transmissible variability. The existence of such a truly wonderful force as it would be was nothing but an invention of the human mind, invented to imagine a way in which species possibly might change.

The species-concept, as we saw, is a concept of absolute constancy ; evolution, however, required inconstancy, and in order to fulfil this requirement variability was invented.

That the kind of variability imagined by Lamarck, at the time when he presented it to the world, was not founded on facts, but was a mere invention only, can not and, as far as I know, has not been denied.

That the variability imagined by Darwin had no better basis, but was simulated partly by diversity within heterogeneous groups, which one generally believed to be homogeneous, partly by segregation of hybrids, is equally certain.

Nor has the mutation of de Vries a better actual basis, since de Vries himself showed that Oenothera Lamarckiana always forms two different kinds of gametes and consequently is a hybrid itself, simulating constancy only, because, of the three possible kinds of combinations of these two gametes, one alone is viable. If we call the two different kinds of gametes, produced by 0. Lamarckiana, A and B, we find that of the three possible combinations of these ; AA, AB, and 88, AB

17

only is viable. Consequently no such thing a 0. Lamarckiann exists . Lamarch'ana is the name of a hybrid, throwing a peculiar kind o s. which action de Vries mistook for a sign of mutation.

We have furthermore seen that Lamarck and Darwin frankly admitted that variability was so slow a process that it could not be experimentally proved, to which Darwin added that its cause was notonly remote bui obscure also, and that de Vries—b} his additional hypothesis of premutation was finally also compelled to accept a unknown, and through its remoteness unverifiable, oause. As a mattei of fact, his species-forming mutation is nothing but a hypothesis of spontaneous generation of genes.

Adherents of : he hypothesis of the exist mce of transmissible variability after de Vries can be brought to three classes. The first of these finds in the word " mutation " an easy term to apply to anything unusual, and seems to obtain a certain amount of satisfaction from the sembla xplanation the application of that term gives to that unusual occurrence. In this way a yellow (lowered Liguslrum found in the dunes of Holland irden, from which it had doubtless escaped, has been described as a iiiuinUo ebbingense of the ordinary Ligustrum vulgare, which is common in the Dutch dunes, and in a similar way an exotic species of Phalloitleac fungi lias been described as a mutation of Phallus caninus. I only mention this cla enthusiastic supporters of the mutation-idea for curiosity sake.

The second of adherents of the transmissible variability conception makes very careful experiments as to the behaviour of organisms, and is satisfied if it finds that organisms which it suj to be pure-bred throw deviations, which it then designates by the name of mutations. The question of the cause of called mutations they for the present leave aside. To this class of investigators belong such ex,•client worker's as Morgan and Baur. of whom non.

The third class of experimenters try to prove thai Lamarck's conception was right— i.e., that changes in the surroundings are the cause of transmissible variability. The chief one of these is Kammerer. These neo-Lamarckists however are no pme Lamarckians, because while Lamarck assumed thai a change in the stimuli, caused by changes in the surroundings alone, could cause a change in organisation, and moreover such a i hange only as favoured the existence of the organism under the new conditions, they are satisfied if they can prove, to their own satisfaction, thai any kind of stimulus causes any kind of change, while they leave the duration of its action—such an important par, of Lamarck's hypothesis entirely aside. Thus has Kammerer. for instance, cul ofi the siphi fan Ascidian. which he kept in aquaria in Vienna, and was satisfied thai he had proved the existence of transmissible variabilit) in the Lamarckian ense, when he had found that iphons of . iv of the.~e ill-treated Ascidians were considerably longer than the siphons of the parents In-fore amputation Not only would the production of a larger organ than the one removed llrious result of amputation one would be inclined to expect the reverse effect, if any but also there was not the slightest proof t hat by producing a longer siphon, the animal adapted itself

18

conditions: it could not even do so, as no change in the aquaria in which they lived had taken placi Moreover the as siphon-. of the Ascidiaus must, during the long time of their existence, havi been injured repeatedly in nature, so th ibly expect that if amputation could cause an effect which was transmitted, the maximum of that effect would long ago have been reached in other words, the siphon of the present time of the wild Ascidian would have the maximum length attainable, li I d thai the amputation had nothing to do with the long siphons produced by the spawn : that the lengthening was simply i I i particular conditions of feeble light and little oxygen in the aquaria to which young animals could respond, while adults could d ' ther word-. the lengthened siphons were mi i nsmiBsible, modifications.

In other cases, however, Kammerer lias indeed Died to influence organisms by a change of conditions. Tin- obstetric toad, Alytes obstetricans, copulates on land and the male carries the eggstrings around its lees until the moment of hatching. The llytes mail no copulation pads on its front legs with which to clasp the female -t frogs and toads which live and copulate in water have. It is supposed that these pads, in the case of llytes have riot been developed because they were no longer necessary, tl of Alytes, by its life on the land, being supposed to be less slippery than females of other genera living in the water Kammerer now kept his Alytes ige, in which they had the choice bi ween land and water, and as he raised the temperature considerably the animals chose the life in the water, copulated there and deposited theii toads and frogs do. In certain in very he got among the progeny males with copulating , I he claims that these were to a certain extent inherited after the animals had again taken to the land. There is a good deal of controversy here. and. to say the least. Kammerer's publication has been incomplete as to controllable data, but if we leave this aside, and accept his unconditionally, do they then prove—as fCammerer claims—an inheritance of acquired characters ? Ido not think they do : they simply show that —as in the case oi /' is modified by changes in its surroundings, the only remarkable thine beinj supposed after effed on the progeny which had again taken to the land. -\. But if this is correct, of which there i- considerable doubt, the result overshoots the mark because normal land-living Alytes havi so that the after effect, if any. can be of short duration only.

it is impossible to speak here, to anj length, about other neoLamarckian experiments; to my way oi thinking n issible effect of changes in conditions has evi i been proved : to me the results of Bonnier with alpine plants an c \» we all know, he showed that halves of perennial alpine individual plants, ' near sea-level, retail of the characters thus obtained, for a ci of years, but that these acquired alpiin unshed completely in the first progeny raised from their

A similar case is the one of Kamnierer with Proteus which proves ihe exact opposite of what hi imagines By hese blind

19

animals living in dark grottos in Austria to the action of the light, he got them to develop eyes. The action of the darkness —during thousands of years—causing the absence of eyes, consequently was not inherited, but lasted no longer than the darkness did.

Neo-Lamarckians will not be satisfied, I fear, by my thus summarily rejecting the results they obtained, and I plead guilty of not doing justice to their efforts by limiting myself to the few cases mentioned ; my only excuse is lack of time and the conviction that, at the present time, the discussion of so-called mutation is of more importance than that of the transmission of acquired characters.

As to mutation, we have to be sure what we understand by this term before we can approach the question whether its existence has been proved or not.

In its modern sense, mutation is a conception of de Vries, so that it is not only fair, but also required by the law of priority, that it should be used in the sense in which he used it.

Of mutations, de Vries recognises two kinds : such as form new species, and such as cause varietal changes only. The first he calls progressive mutations, among the latter he distinguishes between retrogressive and degressive ones, a distinction of subordinate importance. The thing which mutates is, according to his conception, a definite, invisible, living particle of the organism, both the nuclei and cytoplasm being made up of such particles. These invisible particles, these smallest living constituents of the organism, he calls pangenes.

On this basis his definitions are perfectly clear : a progressive mutation is the spontaneous addition of a pangene to the stock already present— i.e., the formation of a new pangene by the organism itself without a previous cross. Such a progressive mutation takes place in two stages, an invisible one in which the new pangene is formed, and a visible one in which it manifests itself by the appearance of a pecies. In this respect de Vries is perfectly clear : On p. 637 vol. 11. of his Mutations-Theorie he says : " Jede progressive Mutation ist im Grunde ein doppelter Vorgang und besteht aus der Bildung einer neuen inneren Anlage (Premutation) und aus der Activirung dieser (sichtbare Mutation) " ; or in English : " Each progressive Mutation is practically a double process ; it consists in the formation of a new pangene (premutation) and in the manifestation itself of that pangene (visible mutation)."

Retrogressive and degressive mutations, on the other hand, are merely changes in pangenes already present, so that he concludes, on p. 644 : " Jede Form welche dureh Neubildung einer inneren Anlage entstanden ist, sollte somit als Art, jede andere welche ihre Eigenthiimlichheit nur einer Umpragung einer bereits vorhandenen Anlage verdankt, sollte als Varietit aufgefasst werden " ; or, in English again : " Consequently each form arisen by the new formation of a pangene should be termed a species ; every other one, which owes its peculiarity merely to the transformation of a pangene already present, should be considered to be a variety onlv."

Between these two great groups of mutations, de Vries thought that the results of hybridisation could decide : On p. 642 hi

20

how: " Retrogressiv und degressh entstandene Formen folgen bei ,\ den entsprechenden Vorfahren den Mendelschen ■.en wahrend entstandenen sich nnisexuell verhalten. ' What this means he states on ... 641 : " Die Bastarde der uuisexuellen Kreuzungen sind constant." So that if we translate thi tences , ; ■• is,. lv and degressive! forms rding to the Mendelian laws, when crossed with their while progressively arisen forms, when so crossed, form constant hybrids."

In order to leave no doubt as to the exclusive importance of progressive imitations in matters of Evolution, de Vries says on p. 714: "Das Product aus der An/.ahl der elementaren Eigenschaften eines Organismus und dem mittleren Zeitintervall zwischen zwei auf einander folgenden Mutationen bei seinen Vorfahren ist der biologischen Zeit gleich. Xennen wir die erstere Grosse M. (die Mutationen), die Liinge der Zeitintervalle L und die biologische Zeit BZ so hal.cn wir also :

M X L = BZ."

It follows from the text, that the term Mutation in this sentence is used in the sense of e mutations, and that the term Biologische Zeit means the time which has passed since the first appearance of life upon our globe, so that we are justified in translating : " The product of the number of elementary properties or pangenes of an organism and the mean interval of time between two successive progressive mutations in its ancestry, equals the time which has elapsed since the Brsi appearance of life upon our globe, which time may be called the biological time. If we call the Brsi value M (the number of progressive mutations), the length of the interval between two si progressive mutations, L, and the biological time BZ, we get the equation :

M X L = BZ."

Let us now suppose that ism has undergone 1000 progressive mutat i I hat the mean interval between two successive mutations lias been 1000 \ would come to the conclusion that the time elapsed between the first appearance of life upon our globe and the appearance of that organism had been one million years.

De Vries therefore a hat the original organisms consisted of one single pangene only and that progressive evolution was due to the gradual addition of new pangenes one by one, by the action of the organism itself, without the aid of crossing.

In this I am unable to see, as no cause is given for the new formation of pangenes, anything else than a hypothesis of gradual creation by means of spontaneous generation of living pangi

Where those pangenes came from remains an insoluble mystery. De Vries tried, however, to find a support for his view that the organism was built up from a number of separate independent particles, by a study of the finer structure of the cell. It was of course well known that plastids, as little as nuclei, were ever formed de novo and were therefore special organoids of the ( ~.|l fined the pangenes to be. De Vries now looked for other such organs and thought he had found

21

them in the vacuoles, which he assumed to be formed by special organoids which he called tonoplasts. and in the hyaloplasm or external layer of the cytoplasm which he imagined to be formed and reproduced in a way independent of the granular plasma. Neither of these views, is, I think, any longer upheld since Pfeffer showed that vacuoles are continually formed de novo.

There was however something else which seemed to give great support to de Vries's ingenious speculations. In order to give a real basis to his view that organisms arc de facto an aggregate of separate living pangenes, it was not necessary to prove thai there was an unlimited number of these in nature : a comparatively small number would be sufficient, as by various combinations of these a practically unlimited number of organisms could be formed, just as the 26 letters of the alphabet are sufficiently numerous to produce a practically unlimited number of books.

Now it seemed at first as if the results obtained by Mendel in his pea crosses gave considerable support to de Vries's conception of the structure of organisms. Mendel, as we all know, had found that the various characters which he had picked out to follow up their transmissions, were independent of one another and could, by various crosses, indeed be combined at will.

De Vries, after the rediscovery of Mendel's work, now assumed that each character was caused by a separate pangene, and this, in connection with Mendel's results, was so plausible that it became generally accepted. Unfortunately for this view, however, it soon turned out that the characters of organisms were not, as Mendel thought, transmitted in complete independence of one another, but in groups, and probably in as many groups as there were chromosomes in the haploid nuclei of the organism, a knowledge which we owe chiefly to Morgan.

In his work Morgan speaks of genes and of mutation, and it has pretty generally been assumed that these genes were synonymous with de Vries's pangenes. while frequently it has been supposed that Morgan's term mutation is synonymous with de Vries's species-forming progressive mutation.

Neither the one nor the other is true. Morgan does not consider his genes as living particles, nor does he assume that genes are ever formed de novo as de Vries does in the case of his new pangenes formed by progressive mutations. Morgan's genes are either chemical molecules or groups of such, whether the one or the other, it is—he says —al present impossible to decide. According to his estimation they are not much larger than the molecule of haemoglobin.

More and more geneticists adhere to the view that what is usually called a gene or a factor is nol an organoid living particle comparable to a pangene of de Vries, but represents a molecule or a group of such in the chromosomes and possibly in the cytoplasm also, although for the present, our knowledge of genes is limited to those present in the chromosomes.

While the term mutation—an equivalenl of transmissible variability—could reasonably be applied to a Living particle suoh as de Vries imagined a pnngcne to be, it ran only cause confusion to apply it to a mere molecule or a group of such.

22

1 -hall therefore in future not speak of mutation of genes, but, using ' non-committal term, speak of alteration of genes, and put the question whether such alterations have been proved in exist, and. if so, by what d !

In order to show the fundamental difference between alterations of genes, considered to be molecules or groups of such, and mutations of pangenes, supposed to be living organoids, 1 may perhaps be permitted to refer t" another view of the constitution of living organisms than that of de Vries, which 1 published in Dutch in 1918.

One <>f the very typical properties of living organisms is their particular shape, so characteristic for each of them, that we distinguish them by these differences in shape, or. to use a Greek term, by their morphological differences. It causes considerable difficulty to imagine a way in which a definite morphos or form could arise by the action of a number of independent particles, orderly arrangement requiring some kind of compulsatory force, directly opposed to independence. There must be. as Blumenbach expressed it already in 1781, a shape-giving force,

Such a force must be a very general one, because, as it has since out, Erankenheim was quite right when he said in 1842 that no such a thing as amorphous substances exist. To cut a long story short, we now know that even each particular kind of atom has its particular shape, caused by the revolving of different numbers of electrons, on circles, on ellipsoids or on courses of another form around a central nucleus, and it is these shapes of the atoms, themselves the constituents of the molecules, which give to the latter their particular shape. Thai 3uch is so indeed could be shown by Debye and Scherrer in the case ol the molecules of benzol, which turned out to be flat discs. Benzol being a fluid— i.e., a substance with little cohesion between the molecules—has, however, not yet a definite shape of its own : it takes, as we all know, just as every other fluid, the shape of an}' hollow object into which we pour it. it has no individual shape, such as organisms have. The most typical individual shapes which we know in lifeless nature, ju racteristic for these objects as the shape of organisms for them, an- crystal--, and we now know through the splendid work of Lauer and the Braggs that such crystals iwe their individual shape to the shapes of the different molecules ami atoms composing them, any particle of a crystal put into an appropriate fluid forcing the molecules and atoms present in this fluid to arrange themselves in the same way as they Mere already arranged in that crystal particle, in which they were so arranged because the shape of the molecules and atoms composing them did not allow of any other May of packing them.

Here then we have something comparable to a germ of a plant, which force- the molecules LI use- a- food to arrange themselves in i he shape of the plant, jusl as the germ of the crystal forced the molecules floating in the liquid to arrange themselves in the form of the crystal ; the secret of the shape of both is a particular molecular arrangement.

Now how far can this analogy be pushed '. The lirst objection which occurs t< bals are hard and living matter is soft, and this—the comparison between an organism and a crystal.

36

put in a general way, is an old one—has always been offered as sufficient to reject it. Since Lehman has taught us that very soft crystals—his semi-fluid crystals—exist, of a consistency quite comparable to that of cytoplasm, this objection has no longer any weight, and one feels the more inclined to accept the analogy as, among Radiolaria, organisms are known of so perfect a crystal shape, that their specific names. octaedrus, dodecaedrus, etc., have been derived from crystal-classification.

Another objection might be seen in the lack of complexity of crystal structure, as compared with that of the structure of cytoplasm. This of course is true, yet, as the Braggs again showed, this structure can in some cases be pretty far advanced, layers consisting of very different molecules succeeding each other in regular arrangements and, what is very important, layers of but one molecule thick, such as physiologists assume to be necessary for many processes in organisms. Water of course is a very essential of life, and the lack of it, or the entirely different way in which it occurs in ordinary crystals, might be offered as an objection, were it not that albuminous crystals are able to imbibe water and swell just as sprouting seeds do, while of course we all know that albuminous matter is the chief constituent of the cytoplasm. Nor are the many inclusions in organisms an objection to their having derived their shape from a crystal-like structure, since we know that, for instance, the Calcite-crystals of Fontainebleau can contain so much sand between their layers that one would assume them to be composed of sand only, and yet they form the typical calcite shape. It seems to me, therefore, that the comparison of the shape of a living organism with that of a crystal can be pushed legitimately so far, that we can say that both owe their shape to a somewhat similar definite molecular arrangement, and, if this be so, we are forced to the conclusion that life need not be a property of each component of the organism—such as de Vries assumed in the case of his pangenes —but may be, and I think it is, a property of the whole, just as many properties, optical, and others—are properties of the crystal and not of each of its components. A living being consequently may be composed of constituents not themselves living.

If my view that organisms owe their properties to definite molecular arrangements be founded on fact, one must be able to show that both the cytoplasm and the chromosomes have a definite molecular arrangement, and zoologists have long ago already shown that some such thing must exist in the cytoplasm of animal eggs, while Morgan’s Drosophila work brought to light a definite molecular arrangement in the chromosomes.

There is however a fundamental difference between a crystal and an organism, to wit, that no chemical reactions occur within a crystal, while a very large number of chemical reactions occur within an organism, which —and this is my greatest objection to those who ascribe alterations within organisms to mutations—are absolutely neglected by mutationists, who operate with their genes as if—mutations excepted—these were the most inert substances imaginable. I will presently have occasion to refer to this important point again.

35

Let us now see what kind of alterations we actually know to occur in organisms, and treat separately of those of which the cause is grosso modo known and of those of which we either admittedly do not know the cause or at the best can suggest one only.

The best known cause of alteration is of course hybridisation, so that we must be perfectly clear as to the meaning to be attached to this term. We do not, as one sees it so often stated, cross individuals, but gametes. This is at once clear when we remember that we can effect a cross also hv self-fertilising a hybrid individual. We. may call this kind of hybridisation haploid hybridisation,. It is not so immediately apparent that we can also effect sub-haploid hybridisation and supra-haploid hybridisation. In the first case one crosses parts of gametes only, in the other case the product of hybridisation contains more than the diploid number of chromosomes. Sub-haploid hybridisation can take place in three ways at least, to wit, by the entrance of one or more chromosomes from the one haploid gamete into the other, by the exchange of an unequal number of chromosomes between those two haploid sets, or by an exchange of parts of chromosomes —so-called crossing-over ; superhaploid hybridisation occurs when nuclei with a higher number of chromosomes than the haploid one unite with one another, or when one of such unites with a haploid one ; the result gives individuals with a higher number of chromosomes than the diploid number, frequently polyploids. Unequal chromosome-distribu-tion analogous to the one occurring in sub-haploid hybridisation may give a complication.

Let us now consider these different cases :

HAPLOID HYBRIDISATION.

Here we have to distinguish between two sub-groups ; in the one the two gametes which unite have an equal number of chromosomes, in the other an unequal one.

The first case applies to peas, with which Mendel got his results, so that we will call it

Mendelian Hybridisation.

Mendelian Hybridisation is a simple affair : a gamete able to produce, when united with another one of the same kind, the character A, crossed with one able to produce with its kind the character a, produces a hybrid, which, whatever its appearance may be, reproduces at maturity the same two kinds of gametes from the union of which it arose. Expressing in equal numbers this simple fact in a formula, without any speculation as to its cause, we get :

A+ a _ s> Aa

which formula, read from left to right, shows us what happens at the formation of the hybrid ; from right to left, what happens when the hybrid forms its gametes. This formula, as Aebly* has recently pointed out, is a formula of reversible chemical action.

*J. Aebly : Ueber die Moglichkeil einer chemischen Deutung tier Bastardir ung und Mendel-spaltung. Vierterjahrschr. Naturf.-Ges. Zurich, 69, 1924, pp. 39-51

38

The basis of Mendelian heredity therefore appear- to i" a of chemical reactions proceeding in one sense in the zygote and in the reverse one at the formation of the gametes.

There must however be a complication, because in an ordinary reversible chemical action, the reaction never' proceeds completely in one sense or in the other, n our case A •< would disintegrate at once after having been formed, into A and a, while the latter two immediately after liberation would again unite to An. This vould go on until an equal balance is reached in which as much A a i- lot med by the combination of A with .... as. on the other hand. A a di-ml into A and a. The factor which determines when this lake- place is the balancing factor K, the value of which changes with temperature, air-pressure, etc

The reaction consequently is incomplete ; to what extent is determined by the value of K. but can. when by reason or other the equal balance is destroyed, become complete, and this will of course happen when one prevents the reaction, reverse to the d one, from taking place.

Now we know that in the somatic cells of the hybrid-, a- a rule no disintegration of A.« into A and « takes place, so that thesi ellmust contain a substance which prevents the reverse reaction from occurring, while this prohibitive substance must be eliminated at the time of the formation of the gametes.

Then, however, another complication must set in. because, if not. the disintegration of A u into A and a would not he completed : a certain percentage of gametes would contain the substance A a. According to Castle this is really what can happen, but the usual belief is that the gametes ate always pure. But the completion of this reaction can very well be explained by the assumption that the substances A and a immediately after formation arc removed so that they can not be combined again, and that disintegration and removal goes on until the last bit of the combination A a has been disintegrated.

It seems to me thai what occurs during the formation of the hybrid and what occurs during the formation of the gametes can very well be reconciled with what we know of cytology. A chromosome A meets in the zygote a chromosome a and the two substances A and a combine to A a and do not enter upon the reverse action on account of a hypothetical substance being formed also which prevents reversion. This substance is removed (luring meiosis. disintegration set- in ami. as, by the moving apart of the paternal chromosomes, the A and a substances are separated, the reaction can he completed. The qui remains whether the supposition that hybridisation and Mendelian segregation are comparable to the two direction- in which a chemical reaction proceeds is a necessary one. and it seems to Aebly that it is. and in this I agree with him. if we assume—as we must —that any kind of substance is the cause of transmissible properties, because then, between such substances, which necessarily in the processes ol life must be such as read readily on one another, a chemical interaction must lake place.

ll' this view approaches the truth, we sec ai onoe whj acquired characters cannot be inherited —they represent nothing but a change

26

m the balance of the reaction caused by external circumstances, which balance is again changed when the external eoudiUons change ; transmissible alterations can never come about in this way.

We see, furthermore, why every individual is likely to be a hybrid in some respe< i ; homozygotes can arise only from two gametes having exactly the same composition, which, considering the complicated chemical react inns which take place in an organism, will hardly ever be the ease, not even with gametes derived from so-called pure lines or even from klonee

The only possible way to prove the existence of transmissible variability would be by means of absolutely pure material, by means of what we are wont to call pmc cultures. This principle of pure cultures hat scellent results throughout; it has shown us that the apparently very variable Draba verna is in reality an aggregate of a large number of constant types, that it was diversity within that 80-cafled species which simulated variability of thai " species." it has shown that the supposedly still greater variability of Algae and Fungi ; which one used to designate polymorphy, was a fairy tale ; it has further shown that within apparently very homogeneous groups of organisms, numerous so-called pure lines can be present. These socalled pure lines, however, are admittedly not pure lines in all respects, but supposed to be pure—l do not say that they are—in respect to one, or a very few characters only ; if one calls certain deviations in respeel to other characters observed within them—as Johannsen did in the case of his so-called pure lines—mutations, instead of considering these unexpected deviations as a certain sign that the line was not yet pure, one practically throws the principle of pure cultures overboard, and opens the door to that old enemy, diversity simulating variability.

I must acknowledge that, with our present knowledge, I see no way to obtain a pure culture of a higher organism. Even if we knew the precise structure of two gametes—which we do not —and if we could unite two gametes of exactly the same kind —which in the case of higher organism- u< cannot, as while both contribute a nucleus to the zygote, one of them only, the egg, contributes the cytoplasm—we would not be sure of obtaining a pure zygote as the two gametes which we unite interact, and we do not know the result of this interaction.

It is better to admit that we are unable to produce what is called a homozygous organism—a term de facto applicable only to an organism consisting of inert genes, and not to one full of interacting substances—than to nourish an illusion which makes us draw unwarrantable conclusions.

We have to face this question : we bring together two gametes of the real nature of which we know practically nothing, which interact in the zygote, then pass through a large number of divisions, making up the soma, of the real nature of which we know practically nothing also.

We may, 1 think, illustrate what we really do, by comparing the two gametes to mixtures of highly complicated chemical substances, and the body to a test tube into which these mixtures are allowed

• Klones are all derivatives from a single individual by asexual propagation

40

to interact. What would one think now of a chemist who, having put into a test tube two substances very imperfectly known to him and who, when the result of their interaction did not come up to his expectations, ascribed the unexpected to a mutation of the elements composing those mixtures ? Yet that is, it seems to me, what mutationists do who operate with genes the nature of which they admit is not known to them.

By this criticism I do not intend to say that we must not use the gene conception in future work on hybridisation : we certainly must while we have nothing else at present to work with, and this way has already given results of great value ; nor do I intend to say that something akin to mutation can not exist—the possibility of the origin of irreversible reactions must be left open —but I do claim that if we speak at the present time of mutation, we speak of it on no better basis than the alchemists spoke of transmutation of the elements, which, though much later proved to exist in some cases, could never have been proved by them, because their knowledge of the elements was entirely insufficient. So is our knowledge of the genes, and I shall therefore from now on leave the possibility of mutation apart and examine what causes we can ascertain for observable alterations in organisms.

In the case of Mendelian hybridisation we know that that cause is the pairing of different chromosomes, not of separate genes, and that these chromosomes must interact, because otherwise it would be impossible to obtain hybrids intermediate between the parents. It is true that one could obtain intermediates also by non-interaction, by mere mixture, just as we can, for instance, obtain something intermediate between the colour of iron and nickel by mixing powders of the two substances, and the separation of the factors —to use a frequently employed quasi-neutral term —could very well be reconciled with such a view, were it not that the so easily reacting albuminous groups which chiefly build up an organism are opposed to such a view, so that the idea of reversible chemical reaction, as explained, fits the case much better.

There is one vexed question to be discussed still. Are we justified to speak of some definite thing—say, some particular group of molecules —causing a character ? It is very easy to say that the blue colour of a flower is caused by the factor for blue ; but is that a fair statement of what happens ? We know nothing about it ; it may or may not, as we can easily see by referring to a non-living substance. Copper sulphate in a crystallised form, in which it contains a certain amount of so-called crystal water, is blue. If we heat it sufficiently to make it lose its crystal water it is blue no longer ; may we now say that the crystal water is the factor for blue? Certainly not, because we know that the blue colour disappears also when we abstract from the copper sulphate either the copper or the sulphur or the oxygen, each of which would therefore have equal right to be considered as factors lor blue, while in reality it is the combined action of all of them which causes the blue colour.

Yet we know that in a good many cases chemists arc abl( that certain groups of molecules in certain chemical substances cause something definite—for instance, a carboxy] group : so that if—as must be the case —a definite molecular arrangement exists in the

28

chromosomes, it may very well be possible to localise certain parts which play a definite role in the production of a certain character, such as Morgan claims to have done, and really seems to have done, in the case of his Drosophilae, of which more anon. But if the ordinary effect ascribed to the localised molecular group fails to appear and is replaced by another similar one, I most decidedly object to consider this as proof of an alteration of that group, because so many factors certainly co-operate or interact in the production of a character, that the change may have just as well occurred in any of them, or even that the change may be due to factors introduced front outside, or by removal of such factors by segregation of a hybrid or even by another internal arrangement of factors already present — such as, for example, takes place in Morgan's cross-overs.

Once more, so-called mutation can not be proved to exist, it seems to me, as long as our knowledge as to what happens in an organism is so scanty as it is.

To resume : Mendelian segregation is the result of an interchange of chromosomes between two gametes possessing an equal number of chromosomes, and the diversity so arising in the progeny depends on the number of chromosomes in the haploid nuclei in which they differ. If this number is n—and we deal with an organism capable of selffertilisation—2 n is the number of possible different combinations. In the pea plant there are 7 chromosomes each of which may differ from its partner in diploid junctions, so that the number of possible genotypically different combinations is 2 7 or 128, all of which were really obtained by Mendel in different crosses.

When we deal with organisms, such as animals or dioecious plants, not capable of self-fertilisation, the number of possible combinations is much larger, as Winge was the first to point out. In such cases, as we know from ourselves, every individual is apt to be heterozygous in one or more characters, hence the diversity among our children. Pairing individuals, of a race not capable of self-fertilisation, therefore usually is pairing two hybrids. Let us now take an extreme case, in which both the male and the female are heterozygous in all chromosomes, and let the haploid chromosome number again be n, then we obtain in Fi already 4 n and in the following generations 10 n genotypically different forms, of which 4 n will be homozygotic throughout. In the case of n being 4, as in Drosophila, we can therefore from a cross of two different individuals obtain 16 different gametes, consequently 32 from the two of them. In Fj such a cross will give already 256 distinct forms, and in F 2 some 10,000 different ones of which 256 are homozygotic. If, therefore, the wild Drosophila flies which Morgan uses in his work, not once but repeatedly even, have not all been homozygotic —which they cannot have been —we need not resort to mutation to explain the origin of aberrant types.

The possible number of forms obtainable from a hybrid non-capable of self-fertilisation increases, of course, much more rapidly than in the case of a self-fertilisable hybrid, with the number of chromosomes which come into play. For instance, Winge has calculated that if we commence by crossing two individuals heterozygotic throughout, of a linneon with 8 chromosomes in the haploid phase, we can obtain 65,536 different types in Fi and one milliard ones in F 2 .

29

After this we need not plead tor the importance of hybrids source of new types.

And that it is really hybridisation which is at the bottom of the pii aiomena which one has called mutation follows, it seems to me. from the expei i' d by Emerson in tl ii /'■'/ hybrids—than whom no one has a greater experience in this field. When 1 saw him three years ago m Ameriea. Emerson told we that he had never across anything which could be explained as mutation in a gamete, but that he had seen such eases in somatic cells — i.e.. where two sets of chromosomes derived from different gametes had met and interaction was consequently possible.

Nor need we doubt thai what has been shown to occur in domestic plants and animals, applies to wild ones too, since it has been shown irould that specimens of the alfalfa-butterfly, taken from nature, at once show deviations—i i . segregations, of exactly the same type as Morgan describes as mutations in his Drosophila cultures.

Tht question is thai a Linnean species, or better, " /.</'<<■ nothing but << group of extremely similar individuals which may product ii large number oj different gametes, which again »'e can see best in the cast oj iln- Linneon, to which wt belong, by il"> diversity oj the /.. <>/' any pair of human beings.

Non-.Memjki.ian Haploid Hybridisation of Gametes with an equal ntjmbbb ob chromosomes.

We have seen that Mendelian Hybridisation depends on an interchange of chromosomes between the hybridised haploid sets at the reduction division Let us now suppose that such an interchange did not take place, what would happen '. Clearly this : we would get precisely the same haploid sets as had made up the hybrids. It is in this way 1 have tried long ago to explain the behaviour of Oenothera Lamarckiana and of all similar Oenothera*. It i> known—nobody, as far as 1 know . denies this—that the great majority of gametes formed by 0. Lamarckiana belongs to two kinds only, which, by Renner. have been (ailed velans and gaudens. On my assumption that no interchange oi chromosomes takes place, the hybrid Oenothera Lamarckiana must on maturity again form these two kinds of gametes, as in fact it does While, furthermore, neither the velans-velans combination nor the gaudens-gaudens combination is viable ; it is only the velans-gaudens combination to which we give the name 0. Lamarckiana, which repeats itself in every generation, and by the death of the two other theoretically possible combinations, simulates constancy ; in fact. 0. Lamarckiana is a self-repeating hybrid. While at the lime u hen I made tin- suggestion there was no cytologic*] evidence for such a view, this now appears to be forthcoming, as it was stated by an American author, in a preliminary communication, that in the ea-e of Oenothera muricata, the chromosomes were so iged in a line that no interchange could take place between them. In other cases irregularities occur, which are the cause of Buch so-called mutations as 0. lata ami others, which have one chromosome more than o. Lamarckiana, while in siill other cases ihe arrangement was such that an occasional interchange of chromosomes oould take place.

30

tn other words, it seems probable that the so-oalled mutations of 0. I.aDinrvkiaan arc due either to an irregular chromosome distribution —of which we shall speak later on—or to an occasional interchange of chromosomes between the two gametes velans and gaudens, which as a rule maintain their individuality, or possibly to crossing over between a velans and a gaudens chromosome.

Non-Mbndeliah Hybridisation between two Gametes with an Unequal Number of Chromosomes.

The first case discovered was that of Drosera obovata, a hybrid between Drosera longifolia and D. rotundifolia , which is common all over Europe. Rosenberg showed that longifolia has haploid 20 chromosomes, rotundi/olia 10, and the hybrid 30. In the pollen mother cells of the hybrid 10 pairs of chromosomes and 10 single ones were observed, so that evidently all chromosomes of rotundifolia (10) had paired with 10 chromosomes of longifolia , while the remaining 10 longifolia chromosomes remained single.

During the reduction division the paired chromosomes were regularly distributed over the daughter nucleus, while the unpaired ones were distributed according to chance. Bach daughter nucleus therefore got 10 chromosomes derived from the pairs and a varying number of single chromosomes. Some of the latter were occasionally extruded into the cytoplasm, where they formed micro-nuclei. In the homotype division all chromosomes were split longitudinally and some were again extruded and formed micronuclei. Fertile pollen never appeared to be formed ; viable ova, to the contrary, were formed, which crossed back with D. longifolia , formed embryos, but could not be induced to ripen, so that no good seeds were obtained.

Still, exceptions seem to occur. While studying hybrids in the Swiss Herbaria, I found specimens collected before they flowered on the 13th July, 1891, by Schroter near Zurich (Katzensee), which he had subsequently cultivated in Sphagnum, after which they had formed normal pollen. Moreover, specimens collected again in the Fallenmoos, near Escholzmett, were evidently segregates, approaching D. longifolia closely, while Schroter again collected specimens of the hybrid near Schuffheim in 1894, where no rotundifolia occurs. They were also collected by Buel in 1919 near Hombrechtikon in pure associations in large numbers while the parents occurred sparingly, while one of the six specimens preserved was again evidently a segregate close to Drosera longifolia. The frequency with which this hybrid is either formed de novo or able to persist, be it by a small number of fertile specimens, is shown by the fact that the records of specimens collected at the Katzensee, near Zurich, start as early as 1841, and that I found the hybrid there still (or again) in June, 1924— i.e., after 83 years!

That forms generally considered to be good species can be formed in this way was first shown, by Rosenberg again, in the case of Hieracium excellens. In the metaphase of this form, the hybrid nature of which was never suspected before Rosenberg revealed it, he found

44

18 pairs of chromosomes and 6 single ones, consequently exactly the same arrangement as was demonstrably caused by hybridisation in the case of Drosera obovata.

This very noteworthy observation, the discovery of the hybrid nature of a form considered to be a pure species, through cytological investigation, was then followed by the astonishing discovery of Tackholm that practically all the wild roses of Europe, North Africa and Western Asia, whose hybrid nature had never been suspected, were hybrids, with paired and unpaired chromosomes, and most of them in all probability Fi hybrids, which maintained themselves since they originated, perhaps thousands of years ago, by apomixis. Some of them however are probably F 2 segregates or segregates of later generations.

While the examples given so far show that hybrids between gametes with unequal chromosome numbers can maintain themselves by apomixis, or occasionally form some normal gametes, Ljungdahl showed in the case of Papaver nudicaule crossed with Papaver striatom, that hybrids between forms differing greatly in chromosome number, the former having but 7 in the haploid stage, the latter 35, can be highly fertile.

The most remarkable thing was that, in this case, no unpaired chromosomes occurred, but that the 42 chromosomes of the hybrid were able to arrange themselves in 21 pairs, so that the reduction division took place in a perfectly normal manner, each gamete getting 21 chromosomes. Consequently chromosomes derived from the same gamete must have paired with one another. Because we know that 7 is the basic number of chromosomes in the genus Papaver, there is no reason to assume that there were homologous chromosomes among the 7 of nudicaule, consequently these 7 will have paired with 7 of slriatocarpum, leaving 28 which among each other have formed 14 pairs. As the nuclei of the pollen contained 21 chromosomes, each of them must, besides a number of slriatocarpum chromosomes, have had a greater or larger number of nudicaule chromosomes, and as in F 2 the reduction division is normal also, they can never after get rid of them.

Consequently the chromosome set of a single gamete can have been derived from a cross between two different so-called species. A haploid gamete consequently can itself be a hybrid, which has derived its chromosomes partly from one, partly from another so-called spv

This is of the greatest possible importance in connection with all work with forms in the gametes of which we see chromosomes of different shape, as this points to their having been derived from different sources.

However this may be, one thing is certain : that the example of the Papaver just quoted proves that even in the gamete we ran not get away from hybridisation, because even the haploid set of chromosomes may have been derived from different souii

SUB-HAPLOID HYBRIDISATION.

Of sub-haploid hybridisation we have just seen an example in the cause of Papaver, in which we crossed parts of gametes only, as we caused a larger or .smaller number of chromosomes from the one

43

gamete to enter into the other one. In this rase we brought this about by bringing the pollen of the one species on the stigma of the other. This is by no means necessary ; sub-haploid hybridisation also occurs, for instance, when we self-fertilise Oenothera Lamarckiana. During meiosis of this peculiar hybrid irregular chromosome distribution takes place, which may cause a chromosome of the one gamete to enter into the other one. In this way two gametes arise, differing in chromosome number, the one having instead of the usual 7 chromosomes 6, and the other 8. These, in the ordinary course of affairs, will be fertilised by non-changed gametes with 7 chromosomes, so that we might expect to get two kinds of individuals, the one with 13 and the other with 15 chromosomes in their body cells. Apparently, however, those with 13 chromosomes are non-viable—such at least have so far not been found—those with 15 however are, and one of these is 0. lata, which de Vries has described as a mutant.

Now let us suppose that 0. lata owes its origin to a velans gamete which has incorporated a gaudens chromosome, then, if we only knew how a velans-velans combination looked, we could determine which changes were brought about by the addition of that single gaudens chromosome.

Something quite comparative to this has now actually been don by Blakeslee in the case of Datura stramonium.

Datura stramonium is a plant, just as certain Oenotheras, frequently found on waste places in different parts of the world, but nowhere really wild, so that its origin is almost equally unknown as that of 0. Lamarckiana. It is said to have come to us as early as the 16th century from the regions around the Caspian Sea, but whether it is truly wild there I have been unable to ascertain.

The chromosomes of Datura are, as those of Drosophila, of different sizes, so that they probably have been derived from different sources. The diploid chromosomes are arranged in 12 sets —in a “ normal ” Datura each set consists of two chromosomes —so that the chromosomal content of a body cell may, diagramatioally, be represented in this way :

Several alterations in the chromosome sets have been observed of which we will, at the present moment, only consider those which came about by sub-haploid hybridisation, due to the taking in of one or more chromosomes of the one gamete into the other one, caused by unequal chromosome distribution, which, as it occurs here, after self-

46

fertilisation, one might also call internal hybridisation. The result will be, for instance, a plant with a chromosome more or less than the normal, say, in the set indicated by A in the above diagram, so that those plants can be diagramatioally indicated in this way, with reference to the A-set, all other sets remaining the same :

The A 1 plant consequently has 23 chromosomes, because it arose from a gamete which had lost one of its chromosomes previously to copulation with a normal gamete, and the A 2 plant has 25 chromosomes, because the gamete which had incorporated the chromosome from its mate copulated with a normal gamete. We can of course indicate the first plant by the formula 2n-I, the latter by the formula 2n -(-1. Of course similar cases arise when the new plants instead of one chromosome have 2 or 3 more than the normal number, with the only difference that the gametes which give off more than one chromosome to their partner appear to be non-viable.

All these cases are characterised by the fact that the sets differ in chromosome number and consequently cannot breed true. Blakeslee called them Unbalanced Chromosomal Types. About them something more should be said ; let us begin with those which lack one chromosome. If we number the sets Ito 12, it is clear that a chromosome may lack in either of the sets, and as it was proved that each of the sets influences the result, or, as it is generally expressed, that each of the sets is the bearer of different hereditary qualities, it is clear that 12 different alterations can arise, depending on the set in which a chromosome is lacking. Of these 12 possible cases, however, so far, but one has been found. This of course has the formula 2n-l.

On the other hand, forms oan arise of the formula 2n 4- 1, in which one set lias one chromosome more than all the other ones. Here also 12 different kinds are possible, all of which have actually been found. Each of these 12 forms consequently owes its characteristics to the fact that the extra chromosome, derived from the partnergamete, is in a different set in each case. These types are most easily distinguished by their capsules, which Blakeslee named : Globe, Poinsettia, Cooklebur, etc.. so that we can also speak of the Globe. Poinsettia, Cocklebur-chromosome, etc.. instead of numbering them. This has some advantages, in the case where in the altered diploids. not 2n | I 20. lint 2n 2 2(1 chromosomes occur, because the two additional chromosomes may occur either in one set or he distributed over two different ones. This, however, can also he expressed in the same manner as before, by applying the formula 2n 2to those eases in which the two additional chromosomes occur both in the same set, and by applying the term 2n - I ■- 1 to the eases where the two additional chromosomes are distributed over two differenl sets

34

Of the possibilities in which 1. 2 or '! chromosomes are involved, the following table gives a resume :

The possibility of an occurrence comparable to such sub-haploid crossing of course is also present, when triploid or quadruploid forms, about which more later on, arise, and the number of possible alterations then increases considerably. Of such, Blakeslee also found a number of oases as the following table shows :

It need hardly be said that many of these forms segregate in a very peculiar way, giving ratios quite different from the normal Mendelian ones, which of course is due to the abnormal arrangement of the chromosomes.

Let us now consider

Crossing-over.

Crossing-over is the exchange of parts of chromosomes. If a chromosome consists, as it appears it docs, of a regularly arranged series of different parts, which we can number, we can represent it in this way :

Chromosome i.

Chromosome if.

1

1

2

2

3

3

4

4

•">

.">

It is clear that when these two chromosomes exchange the parts numbered 3, 4, 5, nothing has been altered : we have still two chromosomes to which the sequence 1, _'. 3, 4. ."> is applicable, But let these

35

chromosomes be different, so that the one contains the parts 1, 2, 3, 4, 5, and the other 0, 7, 8, 9, 10,

Chromosome II

Chromosome I

I

6

2

7

:s

8

9

t

5

10

then an exchange has an effect —we have a real cross between two chromosomes because an exchange of the parts indicated now brings about a different constitution of the chromosomes, which become :

Chromosome I.

Chromosome II

1

0

2

7

3

s

4

!)

10

5

Such a change of constitution of course becomes apparent after segregation and can easily simulate mutation. Morgan himself points to this possibility, where he says, after admitting that there has been found no evidence of any relation between a specific external influence and a particular mutation : " This confession of ignorance* as to external agents causing mutation seems to suggest that the mutation process relates to accidents in the internal processes that take place in the mechanism of cell-division or of maturation or of crossing-over." Moreover he even gives an example which seems to favour such a view. On p. 29 of his last publication! ne savs : " The most significant contribution of Sturtevant and Morgan was to show that reversion of bar—a so-called mutant of the eye, which is the most mutable so-called gene yet discovered in Drosophila —to round (the normal condition) is correlated with crossing-over, at, or very near at the bar locus."

SUPER-HAPLOID HYBRIDISATION.

We now come to the cases where organisms arise with a larger number of chromosomes than the ordinary diploid ones, which of course cannot be sharply separated from the ones just treated.

About the way in which this occurs we have as yet but very imperfect detailed knowledge.

If we leave for the present the distinction between hybridisation and homo-zygotic fertilisation apart, except where hybridisation is evident, we can imagine the following possibilities of an increase of chromosome number

1 by evident hybridisation,

2 by the fusion of more than one haploid nucleus.

3 by a falling out of the reduction division.

4 by a longitudinal splitting of chromosomes.

5 by a fusion in which at least one diploid nucleus takes part

* Morgan. The Genetics of Drosophila. Bibliograpliia Genetica, vol. II 1925. d. 5.

JLtfZO, p. O. f Ibid, p. 29.

36

Of the first possibility, that by evident hybridisation, we have already had an example. You will remember that a cross between Papaver nudicaule with 7 chromosomes in its reproductive cells and Papaver striaiocarpum with 35 gave a hybrid with 42 chromosomes, which arranged themselves into 21 pairs and passed through a normal reduction division. As the basal chromosome number in Papaver is 7, a normal diploid such as P. nudicaule has 14 chromosomes in the somatic cells, while our hybrid has 84; we have consequently obtained a hexaploid form by hybridisation. Many such cases are known in very different families—for instance, in such different ones as Rosaceae and Salioaceae. An imaginary case, in which an increase in chromosome number was caused by hybridisation, was offered by Winge in his hypothesis of indirect chromosome binding.

Suppose two species, each with two chromosomes in their gametes, cross, then the zygote will usually have 4 chromosomes, and, after reduction, form gametes with 2 chromosomes again ; but let the 4 chromosomes of the zygote split longitudinally so that 8 chromosomes arise, then the reduction division will furnish gametes with 4 chromosomes, 2 of which have been derived from each of the 2 parent species, so that a new species has been formed with twice as many chromosomes as the ones originally crossed with one another. Let this new species now cross again with a species with 2 chromosomes, but different from the 2 with which we started, then we will get a zygote with 6 chromosomes which, after splitting, are increased to 12, so that, after reduction, gametes with 6 chromosomes of the last new species arise. In this way series with chromosomes in the proportion 2, 4, 6, etc., which are very common among species of the same genus, may have arisen, but as there are various other ways in which this can come about, we cannot say that they have thus arisen. When Winge offered this hypothesis in explanation of such series of chromosomes, as mentioned, no oytologioal evidence was at all available, nor is there at present for such an occurrence when two species with the same number of chromosomes are crossed.

The principle, however, the splitting of chromosomes, has been shown to occur with certainty in one case at least after crossing two species with different chromosome numbers. I refer to the remarkable results obtained by Bremer in crosses of the sugar-cane Saccharum officinarum with the glagah-grass of Java, Saccharum spontaneum. The former has 40 chromosomes in the egg cells and the latter 56 in its pollen, so that it was to be expected that the hybrid would have 96, while, as a matter of fact, 136 were found. The explanation is that all the officinarum chromosomes had split longitudinally, so that we get in the zygote 80 officinarum chromosomes and 56 spontaneum chromosomes, giving the 136 actually found. Something similar has recently been found by Claussen after crossing two forms of the Linnean species Viola tricolor, which differ in chromosome number : V. tricolor sensu strictu having 13 and V. arvensis having 17 chromosomes in its gametes, but as there are complications in this case, the Saccharum case, so far, is the one most approaching Winge’s hypothesis.

Of the second possible case, increase of chromosome number by the fusion of more than two haploid nuclei, we have to consider whether this is a case of hybridisation or not.

50

The most general case is of course the formation of the endosperm where the two polar nuclei and the sperm-nucleus fuse, and a triploid endosperm arises, but as no case is known where an endosperm gave rise to a new plant, this case does not concern us. It is different with a case described by Ishikawa in 1918, in the Annals of Botany, where he saw an egg cell of Oenothera fuse with two sperm nuclei. As in Oenothera, as we saw, the fusion of the egg cell with one sperm nucleus is hybridisation already, this is a clear case of hybridisation also.

The third case possible, the dropping out of the reduction division, is by itself of no interest, as whether the organism is a hybrid or not, it gives nothing new—it simply furnishes us with a somatic cell capable of reproduction, with a diploid gamete. If this diploid gamete fuses with another one, arisen in the same way, we obtain, it is true, a tetraploid, which however contains nothing essentially new' either, because it only differs quantitatively from the diploid plant in which the reduction division dropped out. The existence of different species with the double chromosome number therefore cannot be explained in this way, but has, as we saw already, to be explained by hybridisation. It is different however when the diploid gamete, arisen by a dropping out of the reduction division, or—to take simultaneously the fourth possible case, by a longitudinal splitting of its chromosomes—fuses with a haploid one. We get then a case of deferred hybridisation, because triploids are not stable, but form gametes with different chromosome numbers. Such a case we may call one of sub-polyploid hybridisation— i.e., hybridisation of part of gametes, the one gamete, as in cases of subhaploid hybridisation, taking in one or more chromosomes of the other one.

The last possible case, the fusion of two diploid nuclei, we have already considered. We saw that this can give nothing essentially new, and we have only to add that this case does occur, not only when the reduction division drops out so that diploid gametes are formed, but that it can also happen that, after wounding, two somatic nuclei in vegetative tissue fuse and give rise to tetraploid plants. Such cases have been described by Winkler in his experiments with chimerae.

We have consequently seen that hybridisation is much more common than was expected—much more common even than at the time when I ventured to offer the theory of hybridisation as a principle of evolution, at which time, for instance, I did not dream of hybrids among mushrooms, which now have been shown to exist and doubtless will be discovered more and more. There is no evidence whatever for progressive mutation in the sense of de Vries, and that for degressive or retrogressive mutation, in a sense somewhat similar to that of de Vries, is very slight also. What has been described as such by Morgan and Baur is much better called alteration, and the cause of it is. m all probability, interaction between gametes /<<>/ alike—in other words. hybridisation. Moreover. I hope to show in the next lecture, that even if it existed it could not be of much importance in evolution, We have furthermore seen that polyploidy can arise by hybridisation. and that, in the few possible oases in which it may conceivably arise without hybridisation, it can produce nothing essentially new,

38

One group of phenomena however still remains to be considered, i he so-called

BUD SPORTS

Bud sports—nectarines on peaches, moss-roses on common roses, to take two examples from Darwin—are, as we know in most cases, nothing but reversion to the two components of a chimera, as, for instance, when Cytisus Adami reverts on the one side to Cytisus Laburnum, and on the other to Cytisus purpureus, from the mere vegetative union of which C. Adami arose.

That chimera-like phenomena can also be present in plants, not arisen by the union of two other ones, has been known a long time, and ascribed by the one to " mutation " of genes, by the majority however, considered as evidence of the possibility of somatic segregation.

Blakeslee has shown that in the case of Datura stramonium, chimeras arisen by changes in the somatic number of chromosomes, give rise to so-called bud sports, so that he speaks of the chromosomal chimeras of Datum.

With certainty, so far, he has been able to establish the following oases :

1. — Chimeras with Chromosome. Deficiencies.

In the summer of 1922, two plants from different lines were found, each with a branch with slight deviations from the normal. The pollen from both of these abnormal branches had considerably more than 50 per cent of abortive grains. Counts of chromosomes in their dividing pollen mother cells demonstrated a deficiency of one of the largest chromosomes which has been shown to be the extra chromosome present in Blakeslee’s (2n -f- 1) alteration known as Rolled.

Offspring from these (2n • 1) branches failed to show individuals of the parental type, a fact which indicates that gametes deficient for the Rolled chromosome are rarely, if ever, capable of functioning.

In the summer of 1923 a single individual was found with a branch similar in appearance and in the degree of pollen abortion to the two chimeras just mentioned, but the failure of grafts to set prevented a count of its chromosomes. Counts of chromosomes in pollen mother cells reveal the cytological condition in the subepidermal tissue only, so that it is possible that these sectorial chimeras were at the same time periclinal chimeras, with an epidermal tissue having a different chromosomal number. That this may have been the case is suggested by the fact that a single plant was found which was markedly abnormal throughout, and distinct from the (2n 1) branches previously investigated, but which was found also to lack one of the same Rolled chromosomes.

2.— Chimeras with Chromosome excess.

A plant, otherwise normal, has been found with one branch bearing leaves and capsules which resembled the (2n + 1) Globe alteration. Chromosome counts had at the time of publication (July, 1924) not yet been secured, but offspring from the normal branches were normal, while offspring from the abnormal branches showed the proportion of Globe seedlings expected from Globe parents. The evidence is clear therefore that the subepidermal tissue of the abnormal branch

52

of the chimera was (2n + 1) with the extra chromosome in the Globe set. That the epidermal tissue was possibly of a different chromosomal constitution is suggested by the fact that neither the leaves nor the capsules on the abnormal branch were fully typical for Globe characters.

3.— Chimeras with double Chromosome numbers

Several cases have been found, chiefly after treatment with cold, in which a single branch on an otherwise normal 2n plant has shown resemblances to a tetraploid. Growth and bud formation in these cases has been poor, but these abnormal branches have been shown to be 4n in generative tissue by the sizes of their pollen grains as well as by the tetraploid offspring which they have produced as contrasted with the 2n offspring produced by the normal branches.

4. — Other Cases.

Other and possibly more complicated chimeras which may have a basis in differences in chromosome numbers are under investigation ; anyhow, to use Blakeslee’s own words : “ The evidence already obtained is sufficient to indicate that chromosomal aberrations may be an important cause in the production of bud sports.” I have only to add that the 2n + 1 and 2n 1 cases mentinoed are of course due to sub-haploid hybridisation, and that the tetraploids are merely quantatively different from the diploid plants.

Let us now see what all this means for evolution ? There is a general agreement that nothing akin to de Vries’s progressive mutation has been shown to exist. On the other hand, both Morgan and Baur claim that changes in genes occur, comparable to de Vries’s degressive and retrogressive mutations, to a certain extent, the cause of which is admittedly unknown. I consider these alterations, as I prefer to call them, as caused by hybridisation also.

Anyhow, the only cause of change which has been definitely proved is hybridisation. We shall therefore consider in the next lecture in how far hybridisation can account for evolution, and also whether such alterations as Baur and Morgan consider as genemutations—i.e., changes in a gene—could be, even if they existed, of any real importance in evolution.

51

LECTURE 111.

The Role of Hybridisation in Evolution.

We have seen in the former lectures that the existence of mutation has not been proved, that Emerson in his extensive work with Zea mais, the Indian corn, states that he has never found anything which could be explained as a mutation in a gamete ; that Morgan admits that the phenomena which he designated by the name of mutations in the case of Drosophila may be due to simple slips in the machinery distributing the genes over the gametes ; that Blakeslee in his no less exhaustive work on Datura, states that he never observed an aberrant form which could be explained on the assumption that it differed in a single gene only from the parent from which it arose. I may now add that Franz v. Wettstein, speaking of the results which he obtained in his excellent experiments with mosses, states distinctly that he never observed anything which could be explained, on the assumption even of hut a change in a gene. Quite recently, however, Baur has expressed other views, which we must now consider a little more in detail. Baur describes a large number of supposed changes in factors, which he calls mutations in the case of Antirrhinum magus. It is not possible to enter into the details of this extensive work, which fills the 4th volume of the Bibliotheca Genetioa, so that I am forced to limit my remarks to some salient points. In the first place, I was struck by the complete absence of cytologioal control, which is the more regrettable since, as we saw, Blakeslee has shown in the case of Datura, that many changes observable in cultures of this plant are not due to mutations, but to irregular chromosome distribution. This absence of cytologioal control is most strongly felt in those cases in which Baur describes mutations in sectors of his Antirrhina, as Blakeslee again has shown that in the case of Datura, chromosomal chimeras exist— i.e., individuals with different numbers of chromosomes in different parts of their body, either in different sectors or in different periclinal layers.

This criticism, however, concerns but part of the mutations described by Baur ; with him I believe that a number of the changes described by him as mutations cannot be explained by irregular chromosome distribution.

We must now first be clear as to the way in which one tries— Baur also—to demonstrate mutations.

This is done in this way. One has believed to be able to demonstrate, in a certain case, by hybrid-analysis, that the individual in question, which in many other respects may be, and certainly is, heterozygous, has a certain factor, say A, in a double dose— i.e., that the organism is homozygous as far as this factor is concerned.

54

A further proof of this result of Mendel analysis we try to obtain by self-fertilising such an individual, and if we rind it constant in respect to the character, which we suppose to be formed by the A factor, we conclude that it really is homozygous in respect to this factor.

Now we repeat the experiment, and find that an individual among the children obtained, or perhaps one among the grandchildren or great grandchildren, which depends on the moment at which we repeated the experiment, is not constant after selfing, but that there are among its progeny individuals which, after back-crossing with their parents, behave in such a way as individuals do to which we ascribe the formula A a.

It is then said that the factor A has mutated to a. I> this interpretation conclusive ? It seems to me not, because in fact, the repetition of the selfing experiment has only shown us that ticplants to which we had ascribed the formula AA are not always constant.

This can have different reasons : we may have overlooked abej rani individuals in our first experiment and therefore have considered the progeny to be homogeneous while it was not. Baur himself say- that he doubtless, in former experiments, has overlooked such small deviations, which he now is able to distinguish, and which he describes as mutations. Morgan assured me that even at the present time, after having seen so many generations of Drosophila, he still finds alterations, which he calls mutations, which he formerly did not recognise. This is of course something which all of us know : an intimate acquaintance with any group of organisms allows us to sec more and more differences ; every one of us is able to recognise different persons among his acquaintances ; the shepherd alone can distinguish the individuals in his flock of sheep, which to us seem all to be alike. The fact, consequently, that one gradually learns to see more and more differences among one's material, may be a cause of serious error in ascribing a mutational origin to a small deviation. But there are other causes of error ; the conclusion that the factor has mutated is based on the belief that the factor was perfectly constant before it became subject to mutation. It was possible, however, that it always was in a labile condition, which found its expression in a different behaviour at different times, now causing an effect such as we ascribe to the action of the factor A, then again an effect as we ascribe to the factor a. Such cases of so-called " reversion " of mutations have already been described by Morgan in Drosophila and by Ikeno in Plantago. Now Baur has also been able to show that in certain cases, factors or—as he, perhaps, more correctly expresses it—chromomcres. show a certain labile oscillation, to which he adds—l translate literally- - "This may appear to many who were convinced of the stability of factors quite a heretical conclusion, yet it is a justifiable one." If this is so, the demonstration of mutations becomes more difficult than it was already, because a mere change in a factor is no mutation, but only such a change as cannot have been caused by crossing— i.e., by a heterozygous condition of the material under investigation. In this respect now, one of Barn's results seems to me to be of prime importance. He himself says thai lie newer saw, in the case of individuals which were homozygous in the factor G. this factor mutate to l;. while, on the other hand, he frequently saw in the case of individuals which were heterozygous in G. changes of this factor to g, and

53

he himself continues : “ Consequently, the very frequent mutation, in an individual Gg of G to g, must be caused in some way or other by its heterozygous condition.” This is quite my view ; without a change of factors, Evolution would be impossible. If one insists to call such changes mutations—a misapplication of the term —the difference of opinion concerns the question whether mutations take place without any apparent cause or as a consequence of crossing, and to me the latter view remains the more logical one of the two. If one keeps in mind how exceedingly difficult it is to obtain absolutely homozygous material, it seems to me that, where Baur himself shows that certain changes depend on a heterozygous condition, on crossing, it is highly probable that the changes brought to light by him in an apparently homozygous material, were merely due to the fact that he mistakenly considered this material to be homozygous. With this view, the fact seems to me to agree that both Baur and Morgan believe to have found several so-called mutations in one and the same chromomere, so that Baur is even forced to assume a chromomere Mutated in a different way for all the different types of variegation observable n the leaves of different Antirrhina.

To me, mutation without demonstrable cause remains a mere word, and admittedly no cause has ever been found, because this term implies that a chromomere is able to remain what it was during a large number of generations and then suddenly becomes able to mutate. For such a change in behaviour there must be a cause, and crossing appears to be the most plausible one, because it is proved that crossing does cause changes.

We know that when a woman with blue eyes marries a man with brown ones, homozygous for that condition, her children will have brown eyes ; the mutation hypothesis assumes, on the other hand, that she would be able to get children with brown eyes if she married a husband of exactly the same constitution as she herself possesses, and moreover that the factor thus changed will itself again remain constant for a large number of generations. This to me seems sorcery.

As Baur considers part of his material in which he observed his so-called mutations to be homozygous, he concludes that pure lines also are able to mutate. Although I grant, as was said in the former lecture, the possibility of an alteration in a so-called pure line, yet in my opinion we will do well to consider deviations of the type in lines which we thought to be pure as evidence of impurity of those lines, because, otherwise, we lose every criterion for purity or impurity. It remains to be said, that Bam’ considers small mutations to be of very frequent occurrence indeed, and that he assumes the differences between different species to be caused by such small mutations, the useful ones of which he, entirely in the sense of Darwin, assumes to have become accumulated by natural selection. This assumption he is forced to make, because he showed that the differences between different types of cultivated Antirrhina are caused by but a small number of different factors, while even those different local races of wild Antirrhina, which resemble one another much more than the domestic races, differ in a large number of factors. To this point I shall refer again.

Although, therefore, I am unable to consider the changes observed by Baur as mutations, I wanted to consider them here, in the first

4a

place, because they certainly are the most exact experiments so far on this field of investigation ; in the second place, because scientific honesty commands to mention in the first place such publications as are contrary to one’s own opinion.* So far as I have been able to study the work, my opinion is that the frequency of the changes observed by Baur is in all probability due to the heterozygous condition of his material and forms a welcome contribution to my view that factors are no fixed organoid particles, but certain groups of atoms, which, by interaction with other ones, can be changed.

To these views one might object, that the hypothesis of the existence of genes has been proved to be a good working hypothesis. I will be the last one to deny it, but so was the conception of the atom, at the time when I learned chemistry, as a round solid globule, while we know now that each atom is a microcosmos, a world by itself. When we remember that but a short time ago the existence of Neon, Argon, Helium, and whatever other names may have been given to the rare gases in the air, which so often had been analysed before, was entirely unknown to us ; when I remember that I was taught that Helium did not occur on earth, and that now gigantic airships are filled with it; when we further keep in mind that but yesterday Radioactivity was even outside the pale of our imagination ; when we remember how difficult it is to get any simple chemical substance—infinitely simple in comparison with the life-substance —in so pure a state that transmutation of it can be demonstrated : we must recognise how infinitely difficult it must be to demonstrate mutation in the living world, even if mutation should exist; at the present time lat least consider this impossible. It is, therefore, it seems to me, unjust to reproach me—as has been done—that I take flight behind possible impurities of the material used, when I oppose mutation. I do not oppose mutation, nor any other form of transmissible variability—l only oppose those who claim that the existence of transmissible variability has been proved ; all I desire is that they prove, to my satisfaction, that they started from material which was so pure, that nothing can happen in it which might simulate mutation. If they are unable to do so, that is neither their fault nor mine, but so long as they are unable to do it, transmissible variability has not been proved, but is nothing than a moans invented to explain evolution.

I fully grant that, theoretically, mutation is possible in impure material also, as is genemtio spontanea in non-sterilised culturemedia ; but nobody will believe the claim that micro-organisms appearing on non-sterilised media have arisen by spontaneous generation, nor can I accept the appearance of sudden changes in not absolutely pure organisms as evidence of mutation. Even in haploid organisms irregular distribution of chromosomes may occur, as Blakeslee has shown in the case of his haploid Daturae.

* I should like to put once more stress on the point, that neither de Vries nor Morgan, nor Baur have caused by the applical to mutate, but thai they i were mutations. Now, hj the machinery of heredity, it is clear thai such a slip can occur at any time, e that the non-ol that the slip did not occur before ii was, upon the experiment, actually observed.

44

If we keep but this in mind ; that in nature at least, two entirely equal gametes in all probability never meet —that there are no pure cultures in nature —we stand, it seems to me, on a firm basis, on which we can try to explain evolution without being hampered by illusions.

We then see that there was no need to invent transmissible variability in order to get a power able to create new forms. Nature has, so to speak, advertised the means it continually uses for that purpose, which it uses in every generation, and from which no higher organism can possibly escape ; sexual reproduction, which, if really two absolutely equal gametes never meet in nature, is nothing but hybridisation. Nature has practically cried out to us : “in each and every generation I give the opportunity for an exchange of qualities by means of this very sexual reproduction and consequently for the origin of new forms ; the diversity of your own children proves it, your gardeners and breeders have applied this principle for centuries and were only able to keep the new forms which they produced in this way by isolating them— i.e., by preventing new crosses. Such limitations I also make use of : by means of isolation in space, or in time — for instance, by differences in time of flowering—or by such absolute means as mutual sterility, but I never eliminate crossing from my course, because by so doing I would out off the way to progression, to evolution. All of you have known the influence of hybridisation, but every one of you has always explained its influence away. But one among you, Kerner von Marilaun, has clearly seen that new forms do not arise by the stimuli of the surroundings, but through the influence of sexual reproduction, and even he was not able to direct himself entirely from universally accepted conceptions, so that he says that, after all, the sperm might possibly vary under the influence of external conditions.”

Among all controversies one fact, and one fact only, is firmly established ; Hybridisation is able to create, simultaneously, a large number of new forms. The exact way in which this happens is still but very imperfectly known ; this to study is the task of the student of heredity—to him, who studies evolution, the mere fact suffices, that from one cross a whole swarm of very different, new forms can arise. In how’ Tar* can' “fE£“ ladl"9s)faiM , "hvt)hrtton 1 ' ■ | ll&a;t’ t iB k 'lfie' qMMh• which we should try to answer.

You see here* a large swarm of very different forms arisen from the cross of two Linneons, partly from Antirrhinum majus by glutinosum and partly from Antirrhinum majus by sempervirens, a linneon very close to glutinosum. From such a swarm one can breed, by isolation, a number of different forms ; one of those, which differs most from its parents, I have called Antirrhinum rhinanthoides, which, if it had been found in nature, would doubtless have been described as an excellent species. This particular form had been formed already, however, when the selection commenced, so that the process which I applied can be compared to the selective action of nature in a swarm of different species, no matter whether these are related to one another or not, it merely being diversity with which nature deals, irrespective of its relationship or origin.

*This refers to a large number of coloured, unpublished, drawings of segregates of the crosses mentioned, which were demonstrated by the Lecturer.

58

A very ciear case of such selection —and 1 believe that the clearest eases are the best—l had an opportunity to see in the neighbourhood of Chicago, where the moving sand dunes of Lake Michigan bury all kinds of trees and shrubs. Of these only those kinds survive which are able to form roots from their stems, because, if they arc not, the sand cuts off the ordinary roots from the necessary supply of oxygen and the plant dies. With this property relationship has nothing whatever to do. Salix humilis, Prunus virginiana, Celastrus scandens, Vitis labrusca, to mention but a few, are thus enabled to keep their leafy branches above the sand ; they grow up with the dune.

hi such cases one usually says that the plants have adapted themselves to the new circumstances, and in this particular case we even know exactly the cause—the moving sand—which has made the trees and shrubs develop roots on their stems ; Lamarckians might say that they adapted themselves directly at the change in conditions of life. A Darwinian, on the other hand, investigating the flora of these high sandwalls, will assume that the species which now send out loots from their stems there, while they do not do so in the nearest proximity of those dunes, have varied in all directions, but that those only which varied in such a way that they could form roots from their stems survived.

As a matter of fact, no variation whatever took place. Neither Lamarckian variability, which requires a long time for the proper development of the new property, nor Darwin’s variability, which can have effect only, after selection through a number of generations, would have been able to save the buried trees and shrubs ; it was too late for them to vary, even if they could. Only those which already were able to form stem-roots, before the moving sand ever reached them, could survive the calamity, just as it is too late for a drowning man to acquire the art of swimming.

Only that can be selected that has already been developed, Vilmorin recognised this, when he said, “ In order to be able to breed something new, we must have it already.”

He also took no account of crossing, and de Vries adapted this statement to evolution, when he said : “ New species are born readymade,” and although this is an exaggeration, because new forms arisen from a cross may for a number of generations continue to segregate before they have attained any degree of constancy, there certainly is a nucleus of truth in it, while de Vries’s opinion, moreover, is very different from Baur’s conception of an accumulation of appropriate small mutations through natural selection.

It is certainly true that the influence of adaptation has been greatly exaggerated. That this is not always recognised is due to the fact that we are but rarely able to find out the cause of a change in nature so clearly as in that of the formation of roots on the stems of plants buried by moving dunes near Chicago. We are only too easily deceived. When we observe how two such different plants as Cacti and Euphorbiae resemble each other greatly in desert-regions, we ascribe this great similarity to the influence of the desert-climate and see in it a support for the view that external conditions cause plants to vary.

46

The similarity, however, is a very superficial one, not much greater than that between an eel and an Ascaris, and certainly not greater than between an eel and a slowwonn.

What we—mirabile dictu—overlook oompletely is. thai in everj other respect the Cactus remained a Cactus, the Euphorbia a Euphorbia so that all their' essential differences have not been affected by tin 1 desert-climate at all. What we overlook also, is the fact that among leaf-forming Cactaceae. in the case of the PereaMae, as well as many leaf-forming Euphorbiaceae, succulent stems were already present, so that it is quite possible that in each of these families crossing gave rise to the formation of leafless forms, which afterwards found their way into the desert, and that we now mistakenly ascribe thai leafless character, which they possessed already before they reached the desert, to the influence of that desert.

That leafless Cactaceae are able to live in very moist regions is proved by some Rhipsalis species run wild in Java; which live there as epiphytes, for instance, as I myself saw' them, on fig-trees. If the climate of Java changed from its present very moist condition to a dry desert-like one, these Rhipsalis species might survive, while the greater part of the other vegetation would succumb, in which case the Rhipsalis would be considered to be a product of the desert climate. That losses of organs, similar to the one here assumed in the case of Cactaceae and Euphorbiaceae, actually do occur as a consequence of crossing, is proved by the fact that among the segregates of my Antirrhinum crosses, plants arose unable to form flowers, and which moreover were succulent while none of their parents were succulent, so that they simultaneously prove that new characters can arise by crossing.

It is true that individuals adapt themselves to changes in the conditions in so far as each individual is the result of its constitution and of the external conditions acting upon it, is, so to speak, a compromise ; but it seems to me that both Lamarck and Darwin made the mistake of considering such adaptations to be transmissible. Fortunately for all of us they are not. Numerous of our ancestors have doubtless fractured some bone in their body, which adapted itself by forming a new kind of bony structure, but all of us inherited normally built bones, together with the power to change their structure to some extent, in accordance with the changed conditions caused by fracture, but we did not inherit the structure already changed. When we admire adaptations we must not forget that we admire extreme cases, and that the same plant, brought under other conditions, is much less astonishing in aspect than the one w 7 e admired so much. The Edelweiss of the Alps or of the dry regions of Siberia cultivated in Holland or England is not nearly so beautiful, not nearly so white and woolly, as in its natural surroundings ; in the matter of adaptations nature deceives us for 50 per cent, at least. Errera, some years ago, sowed the seeds of a Npthoidmax in Brussels and the diversity among the seedlings was simply enormous. This at that time was considered to be an extreme case of variability, while we now of course know that it was segregation of a hybrid. When, therefore, Diels, in the case of the Proteaceae Dyandra, points out that the wideness of the leaf diminishes among the species of this genus with an increase of dryness, this does not prove at all that the leaf shape in this genus is controlled

60

—has been caused by—that dryness ;it may be —and this to me seems a much more likely explanation —that from a very diverse swarm, with different shapes of leaf —such as resulted, for instance, from the Antirrhinum cross —there survived at spots with a different degree of humidity those forms, the leaf shape of which was most in accordance with those conditions —if, at least, the leaf shape has anything to do with their survival. In this respect we must be very careful indeed. We all know that plants of the heath and moorland, such as CaUuna and Erica*, have a form and structure that suggested to some of the earlier workers, and notably to Schimper, that they must be adapted to reduce water-loss from transpiration—that they are, in one word, xerophytes. The peculiar fact that they grew all the same in, apparently wet, moorland was explained by Schimpor by assuming that these moorland soils were physiologically dry, a view which, although it is based almost entirely on the appearance of the vegetation and not put to the test of the experiment, has since been handed on from textbook to text-book. Now Montfort and Hocker have shown that the physiological dryness of moorland soils is a myth, and moreover that CaUuna is not a xerophytc at all. It is true that the single leaf of the ericoid plant may be interpreted as xeromorphic in structure, but a calculation of the total leaf surface per unit of root system puts CaUuna ahead of many mesophytes in its proportion of transpiration surface, and an examination of the amount of transpiration of the plant as compared with its absorbing system, shows CaUuna to be better classed as a xeromorph mesophyte, able to lose water like a mesophyte because in its natural habitat plenty of water is practically always available. What then is the real cause of the characteristic heath and moorland vegetation % Dr. F. E. Clements has recently shown in his monograph upon " Aeration and Aircontent," that one of the governing factors in the peaty soil of the moorland and the heath is the lack of proper aeration in this soil, which renders it unsuitable for the growth of the root-system of a normal plant. Priestley has pointed out that the metabolism of the root-system of different plants may be expected to differ, and that some plants may require less oxygen for their growth ; that the plants on peat are perhaps such plants ; and that, associated with this peculiarity, they have another—namely, that they form unusually large quantities of fat in the roots as they grow. These fatty substances are then sent up into the shoot as an early and abnormally thick cuticle, and then within the shoot again in early deposits of secondary endodermis, and then of cork layers within that endodermis. It was pointed out that these fatty deposits are responsible for certain structural changes, and a later paper, jointly with Miss Beatrice Lee, strengthened Priestley's conviction that a thick cuticle deposited at an early stage will profoundly modify the structure of the young shoot and of the leaves upon which it is deposited. This developmental factor may in the end prove to have a great deal of influence upon the characteristic form of the plants growing on the peat. In this case their characteristic habit is not traced to an adaptation to control water loss, but is found to be a natural developmental consequence of the characteristic metabolism of a root system growing in a soil that lacks sufficient aeration. Wherever the peat-

* This pari' Ij.is been borrowed From an article by J. H. Priestly entitled "Ecology of Moorland Plants" in " Nature," November S. 1924, p. 698.

48

plants grow, whether in peat or in other soils, they will retain this characteristic metabolism, accumulate fats and show the same characteristic structure. But in the peat-soils they reign supreme, because other plants which have not this metabolism fail to grow there.*

The characteristic moorland and heather yegeiatifln consequently has not arisen by slow adaptation, but the properties which enabled the plants which now grow there to lead that kind of life, were already their own before they got there ; they simply found there a soil, unsuitable to other plants, which they occupied. It is anyhow difficult to imagine how slow adaptation could have any effect. How should the well-known window plant, Mesembryanthemum rhopalophyllum, have gradually developed this window and the peculiar leaf structure associated with it, and why. in a country with but little competition and where side by side with it numerous other peculiar Mesembryanthemum species are able to live ? It is exactly this great diversity of Mesembryanthemum forms which grow on the South African veldt, which on me again makes the impression of swarms arisen by crossing. What will now, to return to the swarm arisen from our Antirrhinum cross, become of it if it had arisen in nature ?

If such a swarm finds new soil, without vegetation—for instance, if it can settle on a new railway-dyke—nothing will probably happen at first ; its members will cross-fertilise and the original diversity will, approximately, be retained. In the case of so diverse a swarm as arose from the orossing of the Antirrhinums, this could hardly be tested ; this can only, in a way, be tried in the case of swarms in which very conspicuous and resistant forms are present in small numbers, so that by their reappearance in successive generations it can be tested whether the swarm, in respect to them at least, remains constant. This opportunity arose from the results obtained by crossing Nicotiana a with N. paniculata, which gave, in small numbers, in mutationlike fractions of a percentage, dwarfs and giants. It was shown now, not only in my experiments, that these reappear in successive generations in which the members were allowed freely to cross-fertilise among each other, but also that similar giants and dwarfs appeared in East's a in North America, notwithstanding the fact that he used another form of the linneon—2\ rustiea —than I did. In nature, however, elimination will sooner or later occur in such a swarm, and when, By large lacunae have arisen to separate distinct groups of similar individuals from other such groups, the taxonomist makes of them his species and varieties.

Finally then' may remain, in a certain region, from such a swarm, but one group of so similar individuals that it is considered by taxonomists to constitute a single species. Such a group I was able to obtain, by isolation, from a cross of pumpkins. Within such a group, a certain diversity, however small, probably always remains, because an entirely homogeneous group would be a pure culture, the existence of which I doubt in nature. In order to test this, I induced Dr. Sirks to investigate Chrysanthemum leucanihemum in Holland, where no other linneon of the genus Chrysanthemum, with the exception

peat plants are unable to grow in soil rich in calcium is due to tile hat calcium salts produce insoluble soaps with fatty acids and so plug up the root system of these plants which have so much fat in their roots.

49

of I'hr. segetum in the eastern part of the country, exists, no hybrids of the two being known. The result was that there was considerable diversity within this group, which was generally considered to be a very homogeneous species. There is hardly a group of plants in which so minute differences have been considered to be sufficient to distinguish one species from the other as in the mosses, yet Franz von Wettstein was able to distinguish within the linneon, Funaria hygrometiea, a number of different genotypes, which after artificial crossing tinned out to segregate in the mendelian way.

It was. as we know, the diversity observed within the groups which one took to be a species, which caused the invention of some kind of transmissible variability. As long as one imagined the species to have been created, no other view, indeed, was possible, because one had to assume that the beginning of each species consisted of a single homozygous individual or of a pair of such, and the subsequent appearance of diversity among its or their progeny could only be due to some kind of transmissible variability. How much easier, however, can this diversity within the groups, which taxonomists still insist to call species, be explained if we assume them to be the remains of a once much more diverse swarm arisen as a consequence of a cross, a remnant which can even be reduced finally to a single Jordanon.

To my way of thinking, this is so evident that it seems to me that this very existence of diversity within each group, which taxonomists call species, forms a considerable support for the theory of evolution by means of hybridisation.

1 have heard it objected that crossing happens so rarely in nature. that one cannot base evolution on its occurrence. I believe that this objection is unfounded, that one underestimates its frequency both in the past and at present. The swarm like appearance of large groups of closely related species in certain geological periods, especially among animals living in the sea. such as Brachiopods, Trilobites, etc., could easily be explained on the basis of the theory of hybridisation, the easier, as in the sea a large number of different gametes are liberated, because Loeb has shown that small differences in the chemical compossition of the sea-water—yes. even in its concentration only—suffice tn give foreign spermatozoa entrance to eggs which otherwise refuse to admit them, so that such changes favour hybridisation—a point of importance, because the change in the conditions thus indirectly may bring about the arising of new forms from which the new conditions may make their choice.

On the land, primarily the wind, which carried pollen over great distances, gave plenty of opportunity for crossing, which, secondarily. was considerably increased when insects put in their appearance, and the almost sudden simultaneous appearance of very diverse forms of flowering plants in the history of the earth is in full accordance with these increased possibilities of cross-fertilisation. The book oi the past however will always contain so many imperfect leaves that we will have to complete the text by our imagination : let us therefore see what is known about the frequency of hybridisation at the present time. A looking-over Swiss herbaria gave the following provisional results : The flora of Switzerland contains, as far as Phanerogams and Pteridophytes are concerned. i'(i!i."> species in the sense in which this

50

term is used by taxonomists, As 331 "f the Swiss genera contain but one species each, so thai within these genera no hybrids between species are possible, there remain 2634 species which, theoretically, can hybridise with one another—theoretically only, because many of them never meet. The number "f hybrids found was !*74. or in 42 per cent, of the theoretically possible eases, a very high percentage indeed. In Switzerland there are 693 genera : as 331 of them possesses but one species each, hybrids ran. theoretically again, ocour within ,'!(i2 genera ; such were actually observed within 125 genera— e.g., in 3-1' per rent. of the possible cases

As it was impossible to study, in the short time at my disposal, the actual distribution of the species in Switzerland, and to conclude from this study which species meet and which do not, another test was applied to get an insight in the real possibilities of crossing in Switzerland. It was argued that the more species a genus contained, the better the chances were that some of them would meet so that if hybridisation was a pretty general occurrence, whenever opportunity was offered, the percentage of hybridisation actually occurring should increase in the same rate as the number of species within the genus. This turned out to be really the case.

There are in Switzerland 328 genera with 15 or less species each ; hybrids were found within 101 of these genera, or in 30 per cent, of the theoretically possible cases. 34 genera contain more than 15 species : within these, hybrids were found in 24 genera, or in 70 per cent, of the theoretically possible cases : while there are 17 genera with 22 or more species ; in every one of these latter genera, or in 100 per cent, of the theoretically possible cases, hybrids were found. This agrees perfectly with what Dr. Cockayne said in his recent paper on hybridism in the New Zealand flora : " The main cause of the presence of hybrids, apart from the power of cross-pollination, is the occurrence of related species side by side " : and with the many examples he gives of plants, usually apart, hybridising at once when they meet.

It seems, moreover, to me that a point, also touched upon by Dr. Cockayne, that many species do not “ vary,” as it is called, when isolated, but start varying at once when they meet with another isolated species, speaks greatly in favour of the view that variation, recte diversity, is due to crossing. In Switzerland I saw splendid examples of this great diversity when Rhododendron ferrugineum and Rh. hirsutum, the so-called Alpenrosen, meet. A number of very interesting examples of plants not varying, unless they come into contact w'ith related species, I ow r e to the well-known investigator of the Swiss and related Floras, to Dr. Braun-Blanquet. Two of these may here be mentioned : Saxifraga hirculus shows no variation, no diversity whatever in the whole of Central Europe ; in the centre of the group of species, however, to which it belongs, in Central Asia, it splits up into a number of different forms, Ligularia sibirica (belonging to a genus close to Senecio) is in Auvergne and in the eastern part of the Pyrenees very constant, not showing a trace of variability, while in Siberia, where it meets a number of related species, it forms a swarm of different forms, which can but arbitrarily be differentiated from related species. It seems to me that this does not only show howplants tend to hybridise whenever they get the chance, but also that

64

the fact that they do not vary when isolated, but do when they meet with other kinds, is a fact perfectly analogous to our experience in breeds of domestic animals or in races of garden or other cultivated plants—in other words, that a certain type may be both constant and variable cannot be explained by any theory of spontaneous transmissible variability.

I could give a great many more examples, showing the frequency of hybridisation in nature, but those given now and in the former lectures will suffice, it seems to me.

That crossing as a source of new forms, such as we can observe it in nature, must have its analogon in breeding is self-evident, and numerous examples of it could be given. The clearest cases perhaps are furnished by those cultivated plants which habitually are propagated in an asexual manner. As soon as one sows the grains of such, they show their hybrid origin, as, for instance, our fruit trees, our flower bulbs, etc., which segregate in so complicated a manner that there can be no question of their having originated in any other way. I owe it to the kindness of Bateson, that I can demonstrate here how great a diversity can result from the selfing of a single kind of fruit tree, of the Queen Victoria plum. Sometimes one can see in nature very beautiful examples of hybrids being kept in existence by vegetative multiplication ; I need only remind you of so many hybrid Rubi, multiplying by stolons or by burying the tips of their branches in the soil, of sterile hybrid Potamogetons and Menthae, perfectly able to keep their own by vegetative propagation only—yes, in the case of some such hybrid Potamogetons even being able to supplant the parents. A very beautiful case of such a vegetative propagation of hybrids I saw in Arizona a couple of years ago. In the neighbourhood of Tucson, a linneon of a Candalebre Opuntia, Opuntia versicolor, shows so great a diversity in the colour of its flowers, that there is no doubt of its being a hybrid ; the less so as MacDougal, the Director of the Carnegie Institute there, told me that, according to his observations, all Opuntia species in the neighbourhood of Tucson habitually interbreed. The point which interested me greatly was why this great diversity, the flowers being either white, yellow, or pink, passing into one another and reaching even the deepest wine-red, a diversity reminding one forcibly of the experiment field of a geneticist who happened to have the F 2 of a cross in flower, had been retained in nature. The explanation, to which MacDougall called my attention, was furnished by a great many young plants, arisen from broken-off branches, around many of the adult Opuntiae, a means by which they habitually reproduce themselves.

I An important consequence of the theory of Evolution by means of hybridisation is the possibility of the origin of the same form at different spots of the earth, the so-called polytope origin of species.

Oenothera Lamarckiana arose in all probability in the Botanic Gardens in Paris, where Lamarck brought a large number of Oenothera linneons together : it has never been found wild in nature ; now it forms its segregates at any spot in the world to which it has been transported, at first by the gardener, later by the geneticist. The hybrid Nicotiana rustica X paniculata formed its giants and dwarfsegregates both in my garden in Holland and in that of East in

52

the United States of North America, and nobody will at the present time believe any moro that, for instance, certain white flowered forms of the group of Primula auricula, growing, say, in Austria and in Switzerland, have descended from one another, since we know that they arise wherever Primula auricula and Primula viscosa, from the crossing of which they arose, meet. Hybrids and their segregates, from which taxonomists form their species, arise wherever their parent species meet, just as children of human beings originate where their parents meet.

Hybrids between Europeans and Eastern races are born both in the East and in the West, sometimes even on the way from the East to the West, and so the so-called species of plants and animals, which meet on their wanderings over the earth, may give birth to hybrids at any point where their wanderings intersect. At the Katzensee, near Zurich, for instance, we find the hybrid between Schoenus nigricans and ferruginous, but also in Gotland, in Sweden, near Visberg, and the hybrid between Drosera rotundifolia and longifolia, Drosera obovata, is common all over Europe. No doubt the cross of, say, a Suffolk ram and a Dorset ewe, will give the same results in New Zealand as in England, which by-the-bye could not be expected to be the case if characters did vary according to circumstances. Breeders have long recognised this ; it is their firm conviction that pure breeds do not vary—which makes them pay fancy prices for pedigree stock. It is an important question why, hybridisation in nature being so common as it is, we do not more often observe the formation of new species in nature than we actually do. There is no doubt that of the many segregates of, say, a Cirsium or a Primula cross, or of any other cross een two so-called species, but very little —frequently nothing at all—remains in existence, the longest surviving members being mostly of two kinds only, one of which is most similar to the one, the other to the other parent.

This, however, is just what should be expected on the basis of the theory of hybridisation, because if it is true that the so-called species are the remains of swarms arisen by crossing, nature has been experimenting with crossing for untold aires already, and it is from these experiments that the present-day species have arisen, because they were the most resistant members of such a swarm, arisen by crossing, under present-day conditions.

If we, or if nature, now crosses these most resistant forms which we call species with one another, the result is very likely to be the birth of less resistant forms—under present-day conditions—which then subsequently are again eliminated. It is their superiority under -present-day conditions which gives a considerable amount of stability to the linneon.

If, however, a change in the conditions takes place, the importance of crossing reveals itself, because the new swarms, which j nature's great crossing experiment continues to produce, offer to the new conditions a number of new forms to select from. It is, to my way of thinking, due to this power of offering an abundance of new forms to select from that the chain of life, as Darwin expressed it, has never once been broken ; that life—not a particular form of life—can adapt itself to new conditions.

66

Allow me to refer here, once more, to Baur's most recent results According to him, species have been formed by the accumulation of useful small mutations by natural selection in the sense of Darwin.

On the other hand, he says, literally, that not only after a cross of two wild species of Antirrhinum, in F> an almost unlimited multiplicity of different colours and forms arises, so that even among thousands of individuals hardly two are alike, but that even after a cross of two wild local races of the same species, very much alike, a very complicated segregation also takes place.

This result allows the question why nature, even if mutations should exist, should take the trouble to accumulate these one by one, while in every cross a much larger diversity arises in which the factors in question are already accumulated in its different members. Every cross between two so-called species offers ready-made what, according to Darwin's and Baur's view, she would have to build up laboriously ; every cross offers her for direct selection, as many new forms as she could possibly desire.

And not only new forms—new so-called species—can result from a cross ; old ones, also, even already extinct ones, might possibly arise from a cross, as results from the fact that Love reobtained Triticum forms with a disjointing spindle-rliachis from the cross of two forms which doubtless for many centuries had had no other ancestors than such as possessed non-disjointing spindle-rhaches, from Triticum vulgare by T. turgidum.

The question when, where, and how certain forms arose by crossing, can often be solved by experiments only. I would not be at all surprised if many forms of so-called tertiary age —for instance, among Hieracia —could be reobtained at the present time by crossing, just as easily as I reobtained the green-rimmed Petuniae. which had been lost from cultivation since the middle of the 19th century by crossing the two linneons, from the union of which they arose : Petunia nyctaginiflora and P. riolacea. To plant geographers these points : polytope origin, at different times moreover, and synthesis of the original parents from the segregates of their cross doubtless give a complication to their work, which may not make them feel eager to accept these views.

If from a swarm arisen by the cross of Betxila verrucosa with B. pubescens, such as is found everywhere in Central Europe, where apparently at present the two parent-species do not occur, these or either of them can be reobtained—as they in all probability canthere is no direct connection between these regenerated species and the ones occurring south and north of Central Europe ; possibly it were not they which wandered to Central Europe but their hybrids only. which then have reproduced the parent-species at different spots.

It would be well worth our while, it seems to me. if plant geographers would look into this matter —that is. whether there is any evidence that certain so-called species spread over the earth by means of their hybrid products. This is not at all unthinkable as we now know that hybrids frequently are more vigorous than their parents and able to settle in places which, through their extreme climate, are not accessible to their parents ; many hybrids grow further north or get higher up in the Alps than their parent species, according to uianv authors.

67

The trouble to settle these and many other similar points is that experiments and observations have to go hand in hand to get any definite basis. When I read in a flora, for instance, that of two forms, intermediate between two species, the one is a hybrid and the other a non-hybrid intermediate form, I wonder what led the author to make this distinction. There is a great amount of subjective opinion in this and other similar decisions. We have only to look through different floras, or even but through different editions of the same flora, to see how different the opinions are. In Schinz & Keller’s Flora of Switzerland, for instance, I find in the 4th edition only 23 so-called hybridogenous species, while none of Rosa or of Rubus, for instance, is mentioned, although those genera are doubtless full of them. In one of the former editions Circaea intermedia is mentioned as a hybridogenous species, now as one “ formerly ” considered as such. The time has certainly come that taxonomists, plant geographers, geneticists and oytologists have to collaborate to get a firm foundation for the valuation of different forms. To mention but one example, no taxonomist has ever suspected that by far the greater number of roses growing wild in Europe, Asia Minor, and the North of Africa were not good species, but first-generation hybrids which for countless ages have reproduced themselves by' a process of apomixis, until the cytologist taugThl ’Em so.

We may now consider the question, to what extent crossing can exert its influence. We know that many different kinds of wild Canidae, Jackals, various kinds of wild dogs and wolves can cross with one another, and that from these crosses our different breeds of domestic dogs, such as fox terriers and Newfoundlanders—to mention two extremes—have arisen.

I doubt whether any naturalist is still inclined to believe that these different races have arisen from one another by mutations or any other kind of transmissible variability ; if any of such should still exist he may try to breed Newfoundlanders from Pox Terriers.

It has been shown that in the formation of our domestic poultry at least three different wild species have taken part : different species of pheasants cross easily with one another, and it has also been shown that wild forms of pheasants, described as new species from the Himalayas, had arisen by crossing. One has even obtained, be it sterile, hybrids between guinea fowls and peacocks ; who. then, shall say where the limits of hybridisation are to be found ? Limits there are certainly, some of them drawn by aversion, but to what extent these can be overcome is shown by the conduct of a Llama in the Zoological Gardens at Rotterdam, which as a playmate has a Tapir, and continuously makes efforts to hybridise with it. Males, isolated by any cause whatever from their kind, may thus in nature have played a considerable role in hybridisation, as the human male doubtless has done in his lonely wanderings in countries inhabited by different races.

Limits, however, there doubtless are, although we do not know where to look for them ; are there, for instance, besides nuclear differences, cytoplasmatie differences between various organisms? This we know but very imperfectly ; differences in the structure of the eggs of various animals however point thai way. and the question may

55

here be considered, because it certainly opens interesting points of view. Before considering it we must, however, divest ourselves of the conception that an organism should be an aggregate of separate living particles, such as de Vries imagines his pangenes to be. I prefer to consider it—as was said already in the former lectures—to be something akin to a soft crystal, which possesses both in the nuclei and in the cytoplasm a definite molecular structure, which forces the molecules taken in as nourishment to arrange themselves in a definite manner, corresponding to the particular arrangement of the organism using them as food. In this way we obtain an analogue to morphogenesis in lifeless nature, to that of the crystal which forces the molecules moving in all directions in the mother fluid to arrange themselves in such a way that their motion is limited to oscillations around certain fixed points in the space lattice of the crystal. If we take this view, we can also liberate ourselves from the curious conception that a single gene should be able to cause a single character or a very few of such.

The characters of organisms are just as much the effect of the combined occurrences which happen in the organism and which lead to its particular constitution, as, for instance, the blue colour of a crystal of copper sulphate is due to the molecular arrangement of the different kinds of molecules in that crystal. If I withdraw the crystal water from such a crystal it loses its blue colour. May we now conclude from this that the crystal water is the gene—is the factor—or merely the something which causes the blue colour in question ? Certainly not; for extraction of any of the other constituents of the copper sulphate crystal: of its sulphur, its copper, or its oxygen has the same effect, and I fail to see how it could be otherwise in organisms ; we only imagine we have discovered the particular gene, which causes a particular character, if we have eliminated one of the doubtless many factors which contributed to its formation.

If, on the contrary, we take up the view that the characters of an organ are the result of all that happens in that organism, there is no reason to assume that the chromosomes alone cause the characters—are, as we are wont to express it, the exclusive carriers of the hereditary qualities ; but we must assume that the cytoplasm also plays a role—even a considerable role, probably, in inheritance, as it doubtless does for instance in the transmittance of ohromoplasts in the vegetable kingdom.

Notwithstanding the fact, therefore, that nothing with certainty is known about differences in plasma structure among different organisms, we have to reckon in matters of evolution with a possible influence of cytoplasm, because we need in evolution not only a principle causing change such as we found fully realised in the effects of crossing, but also an element which retains those changes or at least part of them.

All exposers of theories of evolution have recognised the necessity of such a conservative principle ; those who based evolution on transmissible variability also.

I For this purpose Lamarck invented the inheritance of acquired characters, the formation of engrains, as Semon later so clearly expressed it ; Darwin invented the survival of the fittest, by which

56

process the favourable properties were supposed to be accumulated, the unfavourable ones to be eliminated; de Vries invented the spontaneous new formation <>f genes to be incorporated into the set of genes already present.

It seems to me that none of these inventions is necessary. Each and every generation of higher organisms offers not only in the union of two gametes, never exactly alike, the opportunity for an exchange of qualities, but also in the purely maternal transmittance of the ■ cytoplasma. which directly unites all generations with one another, \ the required conservative element.

lf we keep this in mind : that the nuclear fusion offers the opportunity for change, the purely maternal txansmittanco of the cytoplasm, the opportunity for the fixation of changes—wc are, it seems to me, able to form an approximately correct, be it a very rough, of the course of evolution.

As little ground as there is to assume that the groups which wo call species are descended from a single homozygous individual or from a pair of suoh, as little ground there is to assume that but one ldnd of Urplasma has been formed.

It is, on the contrary, much more probable that different kinds of Urplasmata have been formed at different places. Whether these were still without a nucleus, or whether the nucleus, in some form or other, has been present in the very first manifestations of life, is unknown to us, so that we will start our considerations from the moment that organisms with nuclei, more or less similar to the simplest Protista of the present day, existed already.

When these began to multiply sexually, there was not only fusion of nuclei, but fusion of cytoplasm, such as we still find among presentday Flagellates, in Chlamydomonas, for instance ; there was both nuclear crossing and cytoplasmatic crossing, so that the number of different kinds of cytoplasma could be increased just as the number of different kinds of chromosome sets is increased by nuclear crossing.

This period of the possibility of increase of the number of kinds of cytoplasm has, however, not lasted long ; with the differentiation of the gametes into egg cells and spermatozoa the plasma crossing stopped and nuclear crossing alone remained.

The courses of change so became separated, there arose permanently separate cytoplasmatic streams, lines of descent, which were never left by the female nuclei —cytoplasmatic streams which united all successive generations with one another ; and it may very well be entirely due to the existence of these streams that we are able to distinguish large groups of organisms built on a certain plan from other different large groups, the members of which have a particular other building plan in common.

In other words, according to this view, the large Phyla must have « been formed at a very remote period, at the time when cytoplasmatic \ crossing stopped, and this view agrees entirely with the fact, which recent fossil finds impress more and more upon us, that in the vegetable kingdom at least the large phyla originated much earlier in the history n'of the earth than we suspected. Such considerations would lead us

57

therefore to the assumption of the existence of different kinds of cytoplasm in the large phyla at least, and this would explain, for instance, such a curious fact as the one that all skeletons of vertebrates are built up of homologous bones. Changes within such common building plans would then have occurred as ;i i onsequence of nuclear crossing, by which different kinds of chromosomal sets arose, whose action upon the cytoplasm may be assumed to lie able to cause both progressive and retrogressive changes, so that it might also explain the existence of rudimentary organs.

In this way the cytoplasm, by its one-sided maternal transmittance, might become the means of bringing conflicting opinions together, because it might be the substratum in which the phylogenetic history is, to a certain extent, retained, it is true —not, as those who believe in transmissible variability are apt to believe, by means of engrams caused by stimuli emanating from the surroundings, but by the changes caused by crossing.

There is of course a lot of hypothesis in all this, so that I will no longer enlarge upon it, but I should like to insist that the views here exposed are based on two facts : crossing and one-sided maternal inheritance of the cytoplasm, neither of them occurring rarely or in the remote past, but daily and in every generation, so that I cannot help feeling that they offer a firmer basis than any other theory of evolution does.

In as much as we consequently have the two principles which every theory of evolution requires, that of change and that of stability, the first in the crossing, the latter in the one-sided maternal inheritance of the cytoplasm, I think we should test the bearing of these two facts in preference to the mere possibility—transmissible variability—which other theories offer.

Allow me therefore to resume, in a few words, the difference between the views here expressed and those contained in other theories of evolution. The fundamental difference is that the theory of hybridisation ascribes evolution to incorporation within an organism of parts derived from other organisms, while theories based on transmissible variability ascribe evolution to the engrammatio influence of external stimuli.

It was, to my way of thinking, the coming into the world of sexual reproduction, not response to stimuli, which made evolution possible. I may also express this so : that the difference is a double one, in as much as the theory of hybridisation not only substitutes crossing for variability as the cause of evolution, but also considers those groups of individuals which systematists call species as mere remains of formerly very diverse swarms arisen from a cross, and not as the progeny of a single individual or of two individuals which, except in sex, were alike.

All that is needed for the origin of those groups which taxonomists call species is therefore, in my opinion, crossing, which gives diversity, and elimination, which reduces this diversity. As both are known to occur, I consider this view a much simpler one than that of Darwin, which has recently been sharply defined by Baur as the origin of small mutations by unknown forces and the accumulation of the useful ones

71

of these by natural selection. To me it is unthinkable that an accumulation of such small mutations as are expressed, for instance, in a slightly hair covering—an example given by Baur—could play any role in evolution, if we keep in mind how very slight the selective value of so small a difference must be in view of the fact that by far the greatest extermination occurs at a very early age, at the egg stage or seed stage even— i.e., before the characters which might give an individual an advantage over its competitors have been developed. Moreover, we have—as pointed out already—no evidence that it is the fittest, the best adapted, which survives in the struggle for life ; all we can say is that it is the survivals which survive, and this is probably, just as in a railway accident, more determined by a happy position than any particularity of structure, “ Railway accidents,” such as the swallowing of an enormous number of small organisms in one gulp by a whale or in a constant stream of water full of diatoms ingested by an oyster, are very common in nature, and it is hard to see what possible superiority in any respect could save any of these organisms from being ingested ; harder still how this could come about if an accumulation of independently arisen properties were required.

Finally, as I said already, why should nature resort to an elaborate accumulation of properties, arisen one by one, when, so to speak, the finished objects are offered to it, in every cross, in every desirable diversity ? Nature creates by crossing and segregation, and the segregates have to take their chances in which, as little as in human society, it is the superior ones only which succeed.

Man can create nothing, therefore he should not destroy ; the loss of any wild form is to the student of evolution an irreparable calamity, because he knows that it is a loss for ever, and one which may have contained the key to one of the locks of that most complicated problem of evolution, which I have, so inadequately, presented to you in these lectures.

72

PLATES

Mirabilis longifolta

MrRABTLTS JALAPPA X M. LONOWOLTA. One F, plant.

PLATE I.

Mibabilis jalappa

PLATE

Five Plants of the F 2 generation of the cross Mibabius jalappa (female) X M. longifolta (male.), showing the great diversity within this generation. The left-hand figure of the lower row is a branch of the plant shown in the right-hand figure of that row ; it shows irregularly serrate leaves, a character not known to occur in any wild Linneon of Mirabilis.

Mnt.\i;ilis jaiiAppa (female) M. longifolia (male). Flowers of some of the 1. 1 '... plants. Each flower is baken from .1 different individual.

PLATE 111.

PLATE IV.

Miii.MHi.is jiuiTi (female) M. LONG \ (n Leaves of some of the F . plants. Each leaf is taken from a different individual

APPENDIX

LIST OF SUPPOSED WILD NEW ZEALAND HYBRIDS AMONGST THE VASCULAR PLANTS.

BY Dr. L. COCKAYNE

(Those hybridi preceded by a number may be accepted as such with fait confidence.)

FILICES.

Hymenophyllum sanguinolentum x villosum,

1. Hymenophyllum flabellatum x rufescens. Hymenophyllum peltatum x tunbridgense.

2. Adiantum afrine x fulvum.

rt.u.i<»iii liiii (umn urn iiiii. 3. Hypolepis tenuifolia x Dryopteris punctata.

4. Pellaea falcata x rotundifolia.

5. Doodia caudata x media.

6. Asplenium bulbiferum x obtusatum.

7. Asplenium bulbiferum x Lyallii.

8. Asplenium bulbiferum x flaccidum.

9. Asplenium bulbiferum x Hookerianum,

10. Asplenium bulbiferum x Colensoi.

11. Leptopteris hymenophylloides x superba. (There are probably many more fern hybrids but special field studies are much needed.)

TAXACEAE.

12. Podocarpus Hallii X totara. Dacrydium Bidwillii X laxifolium. (Possibly there are crosses involving Podocarpus acutifolius, P. Hallii and P. nivalis.)

GRAMINEAE.

13. Hierochloe Fraseri x redolens. Agrostis filiformis x pilosa.

14. (There are certainly several hybrids in the polymorphic Agrostis fdijormis group which at present is badly understood.)

15. Deyeuxia avenoides x quadriseta.

16. Dichelachne crinita X sciurea.

17. Deschampsia Chapmani X tenella.

18. Danthonia flavescens x Raoulii var. rubra.

19. Danthonia crassiuscula X flavescens.

20. Danthonia pilosa X semiannularis.

21. Danthonia Buchanan! x semiannularis. (The D. semiannularis group contains many jordanons and its composition can only be ascertained by genetic research.)

22. Arundo conspicua X fulvida.

01

23. Koeleria spp. (There are certainly hybrids in the group, but it is possible that Domin's species are themselves hybrids or epharmonic forms.)

24. Poa anceps x caespitosa.

25. Poa anceps x pusilla var. seticulmis.

26. Poa Colensoi X intermedia.

27. Poa Kirkii X MacKayi,

28. Festuca spp. (There are certainly various hybrids between the Ftsfucae but only genetic studies can give reliable results.)

CYPERACEAE.

29. Scirpus aucklandicus x cemuus. (S. inundatus probably consists of species or jordanons and hybrids between them.)

30. Uncinia compacta X divarioata and most likely there are other hybrids in the group into which V. fusco-vagiuuta will come into consideration.

31. Uncinia caespitosa x purpurata.

32. Uncinia caespitosa x riparia.

33. Uncinia caespitosa X rupestris.

34. Uncinia caespitosa X filiiorrnis. And any of the above may also cross.

35. Oarex secta X virgata. Carex inversa X resectans.

3G. Carex Gaudichaudiana X subdola. Carex Gaudichaudiana X ternaria.

37. Carex dissita X Larnbertiana, and probably other crosses in the group, but careful study in the field is required.

JUNCACEAE.

38. Juncus polyanthemos X vaginatus, but the group needs genetic study and field observations.

39. Luzula pumila x ?. (The group of small Luzulae is at present in a state of complete confusion which can only be set right by genetic research.)

40. Luzula campestris though divided into various sub-species is undoubtedly a mixture of many j ordanons or species and hybrids between them, the separating of which would mean many years' research.

41. Luzula ulophylla X one or more forms of campestris.

LILIACEAE.

42. Phormium Colensoi X tenax.

43. Astelia Cockaynei X nervosa var. sylvestris. (Cordyline australis x Banksii is in cultivation.)

ORCHIDACEAE.

44. Thelymitra longifolia probably consists of several jordanons or species and the hybrids between them.

45. Pterostylis australis x Banksii.

46. Pterostylis australis x graminea.

47. Cyrtostylis oblonga x rotund if olia.

FAGACEAE.

48. Nothofagus cliffortioides x fusca.

49. Nothofagus fusca x Solandri.

50. Nothofagus fusca X truncata.

51. Nothofagus Solandri x truncata.

52. Nothofagus ctiffortioides X Solandri.

MORAOEAE.

53. Paratrophis microphylla x opaca.

LORANTHAOEAE

52. Elytranthe flavida x tetrapetala

SANTALACEAE.

65. Mida myrtifolia x salicifolia.

83

POLYGONACEAE.

56. Muehlenbeckia australis x oomplexa.

57. Muehlenbeckia axillaris x oomplexa.

[uehlenbeokia axillaris x ephe H■ muricatella Colenso.

\CULACEAE

59. Clematis hexasephala X indivisa. (Hybrids hs Mr. S. Page from nearly all the other species of 1 i idthej may be expected in nature.)

60. Ranunculus Buchanani X Lyallii. . = /?. Matthewsii Cheeseman. (I have received lately a wonderful senes of hybrids from Mr. J. Scott Thomson, of Dunedin, showing all kmds of transitional forms between the very different parents.)

61. Ranunculus gracilipes X Sinclairii.

w x rivularis

03. Ranunculus macropus X rivularis.

aniifolius X vertieillatus

MAGNOLIACEAE.

H,"). Winters axillaris X colorata.

SAXIFRAGACEAE.

66. Quintinia acutifolia X serrata.

PITTOSPORACEAE.

67. Pittosporum Colensoi X tenuifolium.

68. Pittosporum ellipticum X tenuifolium = P. intermedium T. Kirk.

69. Pittosporum pimeleoides X reflexum.

CUNONIACEAE.

70. Weinmannia racemosa X sylvicoln.

ROSACEAE.

71. Rubus australis X sehmidelioides.

12. Rubus australiß X parvus = R.

ovae-zelandiae X Sanguisorbae var. pusilla.

74. Ai lis X microphylla.

75. Ac ois X Sanguisorbae vars.

7t>. Acaena microphylla X Sanguisorbae yars.

77. Acaena glabra X Sanguisorbae var. pilosa.

Aoaena ovina (exotic Australian) X Sanguisorbae var. pusilla.

LEGUMINOSAE. .11- \y n _..ot.. n «-n

79. Edwardsia microphylla X prostrata.

RUTACEAE.

80. Jlelicope simplex X ternata = M. MonteOii Buchanan.

COBIARIACEAE.

81. Coriaria arborea X sarmentosa.

82. Coriaria lurida X sarmentosa.

83. Coriaria augustifolia X lurida,

ELAEOCARPACEAK.

84. Aristotelia fruticosa X serrata. = A. Colensoi Hook. f.

MALVACEAE.

So. Placianthus betulinus X divarioatus = P. cymoaus T. Kirk.

86. Hoheria augustifolia X Bexstylosa.

63

VIOLACEAE.

87. Melicytus micranthus var. longiusculus x microphyllus.

88. Hymenanthera crassifolia X obovata.

THYMELAEACEAE.

89. Pimelea Gnida X longifolia = P. Qnidia var. pulchella Cheeseman. and many other forms.

90. Pimelea Lyallii X prostrata.

91. Pimelea sp. ( = P. LyalLii as in Cheeseman’s manual but excluding P. Lyallii Hook. f. and P. aridula Cockayne.) X prostrata.

92. Pimelea prostrata X sericei-villosa.

93. Drapetes Dieffenbachii x villosa.

MYRTACEAE.

94. Myrtus bullata X obcordata = M. Ralphii Hook.f. Myrtus obcordata X pedunculata.

ONAGRACEAE.

(There are certainly more hybrids in Epilobium than given below.)

95. Epilobium Billardierianum x junceum (possibly cinereum is the correct name).

96. Epilobium Billardierianum x pubens.

97. Epilobium hirtigerum X junceum (cinereum).

98. Epilobium glabellum X pubenß.

99. Epilobium glabellum X novae-zelandiae.

100. Fuchsia Colensoi x excorticata.

ARALIACEAE.

101. Nothopanax anomalum x simplex. (I have a fine series of forms collected near Lake Manapouri by Mr. \V. A. Thomson, of Dunedin.)

102. Nothopanax arboreum x laetum.

103. Nothopanax arboreum x Pseudopanax crassifolium var. unifoliolatum.

104. Nothopanax Colensoi x simplex (collected by Dr. Lotsy and Mr. F. G. Gibbs near the Dun Mountain track).

(Pseudopanax crassifolium var. unifoliolatum x P. ferox is a possible hybrid recently collected by Mr. C. M. Smith, of the State Forest Service, Invercargill.)

UMBELLIFERAE.

105. Apium filiforme x prostratum

106. (Aciphylla. There are possibly a good many hybrids in this genus, one collected by Mr. W. A. Thomson on the Hunter Mts. may be A. congesta x Cuthbertiana.)

107. Anisotome Haastli x pilifera = A. pilifera (Hook, f.) Cockayne et Laing var. pinnatifida T. Kirk. Anisotome antipoda x lalifolia = A. latifolia var. augustata T. Kirk.

108. Angelica decipiens X montana.

CORNACEAE.

109. Corokia buddleoides X Cotoneaster. = C. Cheesemanii Carse.

ERICACEAE.

110. Gaultheria antipoda x perplexa.

111. Gaultheria antipoda x oppositifolia.

112. Gaultheria oppositifolia x rupestris. = G. fagifolia Hook f. or this may be O. antipoda X oppositifolia.

EPACRIDACEAE.

113. Cyathodes acerosa certainly includes more than one species and between such there are probably many hybrids.

114. Dracophyllum longifolium X montanum.

85

EPACRIDACEAE- ' 'ontinutd.

115. Dracophyllum Lcssoniammi > longifolium.

Gkn (There arc certainly many hybrids between tho species of Gentiana, but much close study in the field is domandod before reliable conclusions can be arrived at. At present it is only misleading to attempt, in most cases, the identification of gentians from single speciniens.)

116. lieiitiana bellodioides x patula.

117. Gentiana bellidioides x divisa.

118. Gentiana corymbifera X ?

APOCYNACEAE.

119. Parsonsia capsularia x heterophylla = P. capsularia var. grandiflora., var. grandifolia, Carse and other forms.

BORRAGINACEAE.

20. Myosotis pygmaea is an aggregate made up of more than one species and many jordanons with hybrids between them, but nothing can be done without field study and genetic research. There are also probably several hybrids between other species of the genus.

SCROPHULARIACEAE.

(Here the generic name Hebe is applied to all the trees and shrubs formerly referred to Veronica.)

Hebe speciosa. (There are many garden hybrids with this as one of the parents.)

121. Hebe Dieffenbachii x Dorrien-Smithii.

122. Hebe macroura X salicifolia var. communis = H. divergens (Cheeseman.) Cockayne

124. Hebe elliptica x salicifolia var. communis = H. amahilis (Cheeseman.) Cockayne and its var. blanda Cheeseman.

125. Hebe elliptica X salicifolia var. communis = //. Lewisii (J. B. Armstrong.) Cockayne but this may be a garden hybrid.

126. Hebe augustifolia X salicifolia var. atkinsonii = x H. Siynmonsii Cockayne.

127. Hebe augustifolia x salicifolia var. communis = graciliima (Cheeseman.) Cockayne, and many other forms. Hebe augustifolia x one or more of the small-leaved species of Hebe.

128. Hebe chathamica is probably made up of more than one species, thoir jordanons and hybrids ; it is an astonishing mixture !

129. Hebe parviflora x ealicifolia var.

130. Hebe leiophylla x salicifolia var. communis = H. Kirki : (J. B. Armstrong.) Cockayne and other forms.

131. Hebe leiophylla X Traversii. Hebe buxifolia X Menziesii.

132. Hebe buxifolia X lycopodioides and probably other whipcord hebes, one of which = H. cassinioides (Petrie) Cockayne.

133. Hebe glaucophylla x Traversii.

134. Hebe glaucophylla x monticola.

135. Hebe monticola X Traversii and probably by other species.

136. Hebe laevis X salicifolia var. = H. Caraei (Petrie) Cockayne, and other forms.

137. Hebe laevis X tetragona.

138. Hebe pimeleoides X salicifolia var. communis = U. Darloni (Petrie) Cockayne and probably other forms.

139. Hebe Buchanani x pinguifolia = H. Buchanani (Hook, f.) Cockayne, var. major, Cheeseman. and other forms.

140. Hebe carnosula X ? gives a great variety of hybrids.

141. Hebe epacridea X Haastii. (There are probably many other species of Hebe which form hybrids.)

>v mv;ii lyjiiii iiy ut. iuo. / 142. Veronica diffusa X Lyallii. Veronica diffusa X lanceolata.

143. Veronica Hookeriana x Lyallii = V. Olseni Colenso.

144. Veronica Lvallii X a whipcord Hebe = Veronica loaanioides J. B. Armstrong.

145. Ourisia caespitosa x glaudulosa = 10. prorepens Petrie in part. (There are probably other Ourisia hybrids but I have not had sufficient opportunity for going into the question in the field.)

65

PLANTAGINACEAi:.

Plantago Brownii x lanigera.

RUBIACEAE.

(There are certainly more Coprosma hybrids than given below.)

146. Coprosma propinqua x robusta —G. Cunnmghamii Hook. f.

147. Coprosma Colensoi X foetidissima.

148. Coprosma Banksii x Colensoi.

149. Coprosma parvifolia x ?.

CAPRIFOLIACEAE.

150. Alseuosmia linariifolia x quercifolia.

151. Alseuosmia Banksii x quercifolia.

152. Alseuosmia Banksii x linariifolia.

CAMPANUL ACE AE.

153. Wahlenbergia gracilis Cheeseman. non A.DC. There are certainly several species included under this name and the hybrids between them.

STYLIDIACEAE

154. Phyllachne clavigera x Colensoi.

COMPOSITAE.

155. Lagenophora petiolata x pumila.

156. Brachycome pinnata X Sinclairii.

Olearia chathamica X semidentata.

157. Olearia angustifolia X Colensoi =Q. Traillii T. Kirk.

158. Olearia arborescens X capillaris.

159. Olearia arborescens , lacunosa.

160. Olearia arborescens X macrodonta.

161. Olearia ilicifolia X macrodonta.

162. Olearia ilicifolia x moschata.

163. Olearia ilicifolia >< lacunosa ~ O. suavis Cheeseman. and other forms.

164. Olearia cymbifolia X nummularifolia.

165. Olearia avieenniaefolia x odorata =O. WUlcoxii Petrie

166. Olearia avieenniaefolia X oleifoiia.

167. Olearia excorticata ..•' ilicifolia.

168. Olearia Haastii ;; moschata.

169. Olearia Haastii x oleifoiia.

170. Olearia (Shawia) coriacea X Forsteri.

171. Celmisia discolor (in a wide sense) x Walked.

172. Celmisia discolor (in a wide sense) X brevifolia.

173. Celmisia densiflora x prorepens.

174. Celmisia incana X intermedia.

175. Celmisia coriacea X Traversii =X @> Morrisonii Cockayne, and other forms.

176. Celmisia coriacea x petiolata.

177. Celmisia coriacea x Lyallii.

Lit. voniiion* uuiiauou A xjy aiiiL. 178. Celmisia coriacea x verbascifolia.

179. Celmisia coriacea X spectabilis = X C. Christensenii Cockayne., C. Botveana Petrie and other forms.

180. Celmisia coriacea X gracilenta.

181. Celmisia Lyallii x spectabilis =C. pseudo-Lyallii (Cheeseman.) Cockayne.

182. Celmisia petiolata x verbascifolia = ('. lanigera Petrie and other forms.

183. Celmisia petiolata x ape i'. moUia Cockayne

184. Celmisia gracilenta x Bpectabilis.

185. Celmisia gracilenta x Monroi.

186. Gnaphalium trinerve x Helichrysum bollidioides.

187. Gnaphalium keriense X Lyallii.

188. Gnaphalium keriense x subrigidum.

zoo. <jruai|jiiciji.LUiii aciicjioo j\ niuiii^iuiiiii. 189. Gnaphalium Mackavi x Traversii.

joy. unapuanum .uacKayi x iraversn. 190. Raoulia apice-nigra x australis.

191. Raoulia australis < lutescens.

102. Helichryaum bellidioidea x prostratum.

193. Helichrysum bollidioides X glomeratum - H. Purdiei Petrie

87

194. Heliohrysum bellidioides ■ Sinolairii = B. Fowerakeri Cookayne.

'miumuu ji. i' ume/u/te/ ( vlii kii \in 195. Hehchrysum coralloides Bel

196. Caaainia fulvida leptophylla.

197. Caaamia fulvida var. Montana Vauvilliersii.

la ■ fulvida var. Mont

-~~. vimnw nunw iui\uui \ MI . UIUUIUniI. ssinia albida Vauvillii

■Wlllin lunula ■, WtUYIIIUMSU. 200. Craspedia. There are several species but not properly defined us vet and hybrids between them.

201. Cotula lanata x plumose 0 i Hook. f.

sericoa.

203. Cotula dioica X pulchella. Cotula obscura x pulchella.

204. Senecio bellidioides Monroi =S. Chriatenaenii Cockayne

205. Senicio Haastii x southlandicus.

206. Seneoio Lyallii x scorzoneroides. (A hybrid between Senecio Hectori and some other species originated in Dr. Hunter’s garden. In Mr. W. A. Thomson’s garden is a hybrid with S. perdicioidea as one of the parents.)

ADDENDA

207. Asplenium flaccidum x Richardi

108. Metrosideros robusta x tomentosa

Printed by ANDREWS BATY a CO. LTD. Hereford Street. Chrutchurch

Please return this item to:

Document Supply Services

National Library of New Zealand

PO Box 1467

Wellington 6140

Supplied at no charge from the collections of the National Library of New Zealand