Art. XXXI.—On Heterostyled Trimorphic Flowers in the New Zealand Fuchsias, with Notes on the Distinctive Characters of the Species.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 25, 1892, Unnumbered Page

Art. XXXI.—On Heterostyled Trimorphic Flowers in the New Zealand Fuchsias, with Notes on the Distinctive Characters of the Species. By T. Kirk, F.L.S. [Read before the Wellington Philosophical Society, 18th January, 1893.] Plate XIX. Exactly a century ago it was observed by Sprengel that the European water-violet (Hottonia palustris, Linn.) produced flowers of different forms on different plants; in one form the pistil was more than twice the length of the pistil of the other. Although convinced that the phenomenon was not accidental, he was unable to discover any reason for its occurrence, and nearly seventy years elapsed before light was thrown on the subject. It is now well known to be a contrivance to insure fertilisation by means of pollen obtained from flowers of another plant, and it has been proved by experiment that the number of perfect seeds in each capsule thus fertilised is much greater than when pollen is applied from the same plant, or from the same form of flower. Pollen from the anthers of either form of flower must be applied to the stigma of the other in order to obtain the most advantageous results. Flowers of this kind are said to be “heterostyled”: but, inasmuch as the difference in the length of the style is often correlated with differences in the length of the stamens, the size of the pollen-grains, and the size of the flower, &c., the late Professor Asa Gray suggested that the term was not sufficiently expressive, and proposed to substitute “heterogone” or “heterogonous”; but, notwithstanding the greater comprehensiveness of the latter term, it has not been generally adopted. Mere difference in the length of the style or stamens, or both, is not of itself sufficient to render a plant heterostyled; there must be a reciprocal relationship between the pollen of one form of flower and the stigma of the other: this may or may not be accompanied by a difference in the pollen and

sometimes in the stigma. On the other hand, heterostyled flowers may not exhibit any great difference in the length of either style or stamens. Hottonia palustris exhibits only two forms of flowers, and therefore belongs to the group of heterostyled dimorphic plants. In Lythrum, Oxalis, and one or two other genera heterostyled trimorphic flowers are developed: in each form there are two sets of stamens, and the style varies in length reciprocally with the stamens, the long-styled form having the shortest stamens, and the short-styled form the longest stamens; one set of stamens in each form is of the same length as a set in one of the other forms. No pistil can be fully fertilised except by pollen from stamens corresponding to the length of its style, but the pistil of either form may be fertilised by pollen from one or both the other forms. The arrangements for cross-fertilisation in the New Zealand species of Fuchsia are much less complicated than in Lythrum or Oxalis, and combine with heterostyled trimorphism a marked tendency to unisexuality, the long-styled form in each species being practically a female flower, although in some instances capable of self-fertilisation.* It it were not for the two forms of hermaphrodite flowers, the New Zealand fuchsias might be termed gynodiœcious. The variation in the form of the flowers of the New Zealand fuchsias has been long recognised, although even yet it can scarcely be said to be fully understood. Sir Joseph Hooker, under the description of Fuchsia excorticata, in the “Flora Nov.-Zel.,” volume i., page 56 (1853), remarks, “The stamens vary much in length, being sometimes quite included”; and was so much impressed by the short-styled form of F. procumbens that he was led to consider it a distinct species, and described it as such under the name of F. kirkii. After pointing out that it was impossible to distinguish the two plants by the leaves alone, he adds, “In the flowers they differ widely…. At first I was inclined to think that these differences might be sexual; but I should rather regard them as diagnostic of two representative species that possibly had a comparatively recent origin.”†† Ic. Pl., 3rd ser., i., 67. Mr. G. M. Thomson detected two forms of flowers, one of which he rightly stated to be hermaphrodite in structure but pistillate in function.T† Trans. N.Z. Inst., vol. xiii. (1880), p. 264. Unfortunately, with other southern botanists, he failed to distinguish between F. colensoi and F. procumbens, and consequently has no distinct reference to the heterostyled flowers of the latter. All the New Zealand species agree in having alternate leaves and axillary solitary flowers, although in F.excorti-

cata the flowers are sometimes aggregated on very short branchlets, with undeveloped internodes. They further agree in having deep-blue dry pollen, bound together with extremely delicate threads, which are slightly viscid,* This character is exhibited by several South American species—probably by all. and penetrate the mass in all directions, so that the pollen-grains readily adhere to the feathers of birds which frequent the trees for the sake of the honey afforded by the flowers.†† The chief agents in effecting fertilisation in F. excorticata, are the tui (Prosthemadera novœ-zealandiœ), the bell-bird (Anthornis melanura), and, in the extreme north, the stitch-bird (Pogonornis cincta). I suspect that the parrakeets Platycercus novœ-zealandiœ and P. auriceps assist in the process; the white-eye (Zosterops lateralis), and in some cases the naturalised sparrow, although not honey-feeders, certainly render assistance, as the blue pollen-grains are frequently found on their feathers: they doubtless frequent the tree in search of insects, while the tui, the bell-bird, and the stitch-bird are attracted by the honey. It will now be convenient to describe the general structure of the different forms of flowers produced by each species. Fuchsia excorticata, L. fil. In this species the flowers are pendulous, and at first of a deep-purple colour blotched with green; ultimately they assume a dull-red tint. Immediately above the ovary the calyx-tube is dilated in a globular form, then suddenly constricted, when it becomes funnel-shaped; the limb is divided into four acute spreading lobes; the tube is marked externally by eight more or less elevated ridges, caused by the insertion of the filaments; the deep violet-coloured petals are very small, and alternate with the segments of the calyx. The introrse anthers are carried on rigid but very delicate filaments, which vary in length even in the same flower, and are attached to the anthers in an obliquely-peltate manner. The stigma is globose, minutely papillose, and obscurely four-lobed on the upper surface. Three principal forms of flower may be observed, but each plant produces a single form only. The different forms may be described as under:— 1. The Long-styled Form.—In this form the flower is slightly smaller than either the mid- or short-styled forms. The style is fully twice the length of the calyx-tube, but the filaments are so extremely short that at first sight the anthers appear to be sessile, and in some instances are partially included in the calyx-tube. The pollen-grains when present are yellow, and almost invariably abortive. The petals are very small, and often contorted. 2. The Mid-styled Form.—The style is much shorter than in the previous form, being about one and a half times the

length of the calyx-tube; the anthers are carried on long filaments, which are shorter than the style and of nearly equal length, although showing a slight tendency to form two sets. The pollen is of a deep mazarine-blue and well developed, the grains differing but slightly from those of the short-styled form. Petals larger than in the long-styled form, never contorted. 3. The Short-styled Form.—In this form the style is shorter than in either of the preceding, while the stamens are longer and of unequal length, the longer alternating with the shorter and nearly equalling the style, or in some cases exceeding it. Petals rather smaller than those of the mid-styled form, but never contorted. Fuchsia colensoi, Hook. f. 1. The Long-styled Form.—This closely resembles the long-styled-form of F. excorticata, but the filaments are rather longer; the pollen is usually abortive. Petals minute. 2. The Mid-styled Form.—This also corresponds to the mid-styled form of F. excorticata, but the filaments are of equal length and more slender. The pollen is of a deep-blue colour. Petals minute. I have not seen the short-styled form of this species, but entertain no doubt of its existence. As a rule this species does not produce flowers so freely as F. excorticata, and my opportunities of examining it during the flowering-season have been comparatively few. Fuchsia procumbens, R. Cunn. The flowers of this species are invariably erect and apetalous; the calyx-tube is neither inflated at the base nor marked by raised longitudinal ridges, while the segments are always refracted and never spreading. The stamens are exserted and of uniform length. 1. The Long-styled Form.—In this form the style greatly exceeds the stamens, and the stigma is distinctly four-lobed; the anthers are small, and the pollen, although of the same deep-blue colour as the perfect pollen in the preceding species, is probably abortive, but the material at my command is not sufficient to allow this point to be determined. Flowers smaller than in the mid- or short-styled forms. I have not seen this form in a growing condition. 2. The Mid-styled Form.—In this form the style exactly equals the stamens, but the flower differs from the preceding in no other particular, except its larger size. 3. The Short-styled Form.—The style of this form is shorter than the calyx-tube, in which it is wholly included. Pollen bright-blue.

Flowers of N. Z. Fuchsias

In F. excorticata the three forms grow intermixed, usually in about the same proportion; although the flowers of each tree are uniform, there is a considerable amount of variation in the flowers on different trees, so that it would not be difficult to find intermediate forms, as in Lythrum grœfferi and L. salicaria. The long-styled form of F. excorticata is practically a female flower, as the anthers are almost invariably abortive, and it is especially worthy of note that it produces a larger quantity of fruit than either the mid- or short-styled forms. As the three forms are usually found together wherever the plant is plentiful, it may fairly be assumed that this profusion of fruit is largely due to the application of pollen from both the mid- and short-styled forms; but this point can only be determined by an exhaustive series of experiments. The assumption receives a certain amount of support from the fact that, in a few observed cases where the mid-styled form appeared to be absent, the quantity of fruit produced by the long-styled form was greatly reduced. The same result has been observed in the absence of the short-styled form, and it may well be that the paucity or entire absence of fruit on many trees, even after flowers have been produced in profusion, is at least partially due to the absence of one form or the other. So far as known to me, F. colensoi produces fruit but sparingly, especially in the North Island. May not this be partially due to the absence or comparative rarity of the short-styled form? No instance is known, so far as I am aware, of the different forms of F. procumbens growing intermixed. I have never seen or heard of more than a single form occurring in any one locality. This may well account for the fact that the handsome fruit of this species has not been seen in the wild state. At Tryphena Bay Professor Hutton and myself examined hundreds of flowers, but saw no trace of fruit; subsequently I had the same experience at Mine Bay, and again at Whangaruru. Its original discoverer saw no trace of fruit at Matauri, where he collected the plant during the autumn months, when fruit should have been plentiful. Although the mid- and short-styled forms are often cultivated, I have never seen both forms in the same garden. At present the long-styled form has not found its way into cultivation in the colony, although it appears to have been cultivated in England for fifty years, and the short-styled form for twenty years. Cultivated plants of the mid- and short-styled forms grown separately, although most frequently sterile, produce fruits occasionally; sometimes only a single berry becomes mature, rarely more than two or three. On one

occasion I saw a specimen in a greenhouse bearing nearly a dozen berries, but this was a solitary instance. The seeds are numerous, but possess a low power of germination. Of the total number of seeds contained in two large fruits from a short-styled plant only two germinated; both came true to the parent, and produced short-styled flowers. The rarity of fruit on cultivated plants and its absence on wild plants is surprising when it is remembered that the flowers are erect, and that the anthers and stigma of the mid-and short-styled forms are in such a position that the transfer of pollen by minute insects could scarcely be avoided; moreover, there is no evident tendency to dichogamy. F. procumbens is an extremely rare and local plant; it is far from abundant in any of the few localities in which it is found. From the absence of fruit in the wild state, in which only a single form of flower is present, and its rare occurrence in cultivation under the same condition, it must be inferred that each form of flower is sterile with its own pollen, or, at least, that it is not capable of fertilisation in any appreciable degree. The same cause doubtless accounts both for the rarity of the plant and its restricted distribution, as well as for its remarkable uniformity in habit and foliage when compared with F. colensoi and F. excorticata. The following conclusions appear to be warranted by what has been already stated:— 1. Self-fertilisation occurs but rarely, and, so far as evidence is available, the number of perfect seeds produced by self-fertilised flowers is extremely small. 2. The long-styled flowers of F. excorticata and F. colensoi are practically female flowers, but produce fruit more abundantly than either of the hermaphrodite forms. 3. The hermaphrodite forms are reciprocally related, and have a special relationship with the long-styled form. The precise amount of advantage derived from the fertilisation of the long-styled form of F. excorticata with pollen from both mid- and short-styled forms can only be determined by a long and careful series of experiments. Some advantages must be obtained by the reduced demands upon the vital energy of the plant arising from the non-development of pollen. As the mid- and short-styled forms of F. procumbens are not unfrequent under cultivation, it would be comparatively easy to ascertain the effects of intercrossing between these forms. The long-styled form is rare, and not easily obtained. It is most desirable that this form should be placed in the hands of the cultivator, since the species is unable to extend itself in a state of nature, and is gradually dying out; while

its extinction is hastened by the progress of settlement. I have been informed that it has already become extinct in Tryphena Bay. It is remarkable that the differential characters of F. colensoi and F. procumbens have not been more fully defined: the original drawing of the long-styled form of F. procumbens was published in 1842, and the short-styled form (F.kirkii) in 1871. In the original “Flora Novæ-Zelandiæ,” volume i., page 57, F. colensoi is evidently included under F. procumbens, although the flowers are said to be apetalous; and in the “Handbook of the New Zealand Flora,” page 76, the flowers of F. procumbens are said to be “as in F. excorticata, but smaller.” At page 728 F. colensoi is for the first time described as a distinct species, but very briefly. The Otago botanists have usually mistaken forms of F. colensoi for F. procumbens, although the erect flowers of the latter are alone sufficient to distinguish it from any other species. I venture, therefore, to give the following amended descriptions of the New Zealand species from the MSS. of the “Students’ Flora of New Zealand,” now in course of preparation:— Fuchsia excorticata, Linné, Supp. 217. A shrub or small tree 10ft.–45ft. high; trunk 6in.–3ft. in diameter, clothed with brown papery bark. Leaves alternate, ½in.–4in. long, lanceolate or ovate-lanceolate, acute or acuminate, entire or obscurely toothed, membranous, silvery beneath; petioles short. Flowers axillary, solitary, ¾in.–lin. long, trimorphic, on filiform drooping peduncles; calyx globose at the base, then suddenly constricted and expanded into a funnel-shaped tube, with longitudinal ridges, segments 4, acuminate, spreading; petals very small; stamens exserted and, like the style, varying in length. Fruit a pendulous purple or black berry. Link et Otto, Abb., t. 46; Lindl. in Bot. Reg., t. 857; D. C., Prodr., iii., 39; A. Cunn., Precurs., n. 533; Raoul, Enum. Pl. Nov. Zel., 49; Hook. f., Fl. N.Z., i., 56; Handbk. N.Z. Fl., 75; T. Kirk, Forest Fl. N.Z., t. 36 and 36A. Skinnera excorticata, R. and G. Forster, Char. Gen., t. 29; A. Rich., Fl. Nov. Zel., 331. Agapanthus calyciflorus, Banks and Sol., MSS. Kotukutuku, konini (the fruit only). Hab. On the margins of woods, c., from the North Cape to Stewart Island. Ascends to 3,000ft. Flowers August to December. In exposed or elevated situations this species is reduced to a dwarf bush. The branchlets are rather stout and very brittle. The petioles and peduncles vary greatly in length.

F. colensoi, Hook, f., Handbk. N.Z. Flora, 728. A small erect or prostrate shrub with slender branchlets. Leaves alternate, ovate or orbicular-ovate, rounded or cordate at the base, acute, very membranous when dry, obscurely toothed, rarely silvery beneath; petioles very slender, longer or shorter than the leaves. Flowers as in F. excorticata, but the calyx-tube rather wider at the mouth, and the petals minute. Hab. Lower Waikato, southward to Stewart Island; more frequent in the South. Ascends to 1,500ft. Flowers October to February. This species is local in many districts, and always less abundant than the preceding; the flowers are never produced in such profusion. In some places the unbranched flexuous shoots are 8ft. or 9ft. long and subscandent; in others the entire plant scarcely exceeds a foot in height. F. procumbens, R. Cunn., MSS. in A. Cunn., Precurs., n. 534. Stems extremely slender, prostrate, 6in.—18in. long. Leaves alternate, rounded - ovate or cordate, obscurely toothed, ¼in.–½in. in length, shorter than the slender petioles. Flowers axillary, solitary, ½in.–¾in. long, on erect peduncles; calyx-tube funnel-shaped, without raised ridges, not dilated at the base; segments oblong, acute, recurved. Petals O. Stamens equal; style varying in length, stigma capitate, four-lobed. Berry large, clavate, glaucous, bright-red. Hook., Ic. Pl., t. 421; Raoul, Enum. Pl. N.Z., 49; Hook. f., Fl. N.Z., i., 57; Handbk. N.Z. Fl., 76. F. kirkii, Hook. f., Ic. Pl., t. 1083. Hab. North Island. In sandy or rocky places near high-water mark; rare and local. Matauri; Whangaruru; Cape Colville Peninsula; Great Barrier Island. November to February. This appears to be the only species with erect flowers, and the only apetalous species with the calyx-tube destitute of external longitudinal ridges. Explanation Of Plate XIX. Fuchsia excorticata, L. f. 1. Long-styled form. 2. Mid-styled form. 3. Short-styled form. Fuchsia colensoi, Hook. f. 4. Long-styled form. 5. Mid-styled form. Fuchsia procumbens, R. Cunnc. 6. Long-styled form. 7. Mid-styled form. 8. Short-styled form. 9. The same in longitudinal section.