Art. V.—The Moas of New Zealand.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 24, 1891, Page 93

Art. V.—The Moas of New Zealand. By Captain F. W. Hutton, F.G.S. [Read before the Philosophical Institute of Canterbury, 1st October and 4th November, 1891.] Plates XV.-XVII. Contents. Introduction. Classification of the moas—  Families.  Genera.  Species.  Sexual differences. Characters of the Dinornithidæ. Explanation of measurements. Table of ratios in the genera. Characters of the genera and species—  Genus Dinornis.   Sub-genus Tylopteryx.   Genus Palapteryx.   Genus Anomalopteryx.   Genus Cela.   Genus Mesopteryx.   Genus Syornis.   Genus Euryapteryx. History of the moas—   Origin of the Ratitæ.   Advent of the moas in New Zealand.   Development of the moas.   Extinction of the moas. Table of measurements of the species. Introduction. Nothing connected with the natural history of New Zealand has attracted more attention than the extinct moas. Ever since 1840, when Sir R. Owen announced to the scientific world the former existence in New Zealand of a large struthious bird, exceeding the ostrich in size, interest in the subject has not flagged. And the discovery that the moas were wingless, and had formerly existed in great numbers and in great variety, opened up several new problems for naturalists to solve. With the exception of the moas, every known bird, whether living or dead, has wings. In some they are so small as to be useless, as in the kiwi and the cassowary; while in the penguins they are used only for swimming. Still, some kind of fore-limb exists in all birds but the moas. We know that the moas were wingless, partly by the negative evidence of no wing-bones having ever been found, and partly by the positive evidence of the skeleton, which has no cavity for the articulation of a wing, and in many cases does not show the existence of any shoulder-girdle at all. However, as will presently appear, it is probable that one genus of moas (Palapteryx) had rudimentary wings like the kiwi. Our knowledge of the moas, of their structure and of their history, is very fragmentary, and the information is scattered through various publications. There is considerable confusion in the characters given of the different species, and different opinions are held as to when the moas came to New Zealand and when they became extinct. Under these circumstances I have thought that an attempt to clear up the confusion and to give a connected account of what is known on the subject would prove useful as a basis for further investigation. With

this object in view I have been collecting materials for some time. I have examined most of the collections in New Zealand, and I have collated all the measurements hitherto published—those by Sir R. Owen in the Transactions of the Zoological Society; those by Sir James Hector in the Proceedings of the Zoological Society for 1865 (p. 751); as well as those given in various volumes of the Transactions of the New Zealand Institute by Sir Julius von Haast, Mr. Thorne, Mr. Taylor White, and by myself. Mr. Aug. Hamilton has also supplied me with measurements of all the leg-bones found in the Patangata Swamp, near Te Aute—a large and valuable collection which has been most useful to me. I have also to thank Sir James Hector, Professor T. J. Parker, Mr. H. O. Forbes, Mr. T. F. Cheeseman, and Mr. R. J. Kingsley, for allowing me to examine the collections of moa-bones in the Museums of Wellington, Dunedin, Christchurch, Auckland, and Nelson respectively. In addition, I have received valuable information from private collectors, including Mr. F. W. Stubbs; Mr. Mitchell, of Manapouri Station; Mr. W. Colenso; and the Hon. W. Mantell; to all of whom I here tender my best thanks. The magnificent series of plates published by Sir R. Owen in the Transactions of the Zoological Society, and reissued in his “Extinct Birds of New Zealand,” have made the collections in London known to all naturalists, and they have been of the greatest use to me. In my paper I have quoted these plates from the “Extinct Birds,” and not from the Transactions of the Zoological Society, because the former is to be found in all good libraries in New Zealand, while the latter are very rare. By the means just enumerated I have become possessed of an amount of information far greater than has been got together before, and it has, I think, enabled me to clear up the classification of the moas, as well as to throw some new light on their distribution and history. My work is founded on the measurement of the leg-bones of individual birds belonging to sixteen different species, and from these I have inferred with considerable certainty the proportions of the leg-bones in the other species. There is still, however, in many cases much conjecture in placing the other bones of the skeleton with the legs. In those genera of which I have seen a considerable number of skulls, sterna, and pelves, the chance of error is small in assigning the largest bones to the largest legs, and vice versâ, and then fitting in the intermediate forms. And when we have an occasional check in an individual skeleton the chances of error are further reduced. Again, geographical distribution sometimes gives a clue to the assignment of bones, some species being found in both Islands, and others only in one. And, lastly, we have relative numbers to help us, as the

commonest skulls, &c., will usually belong to the commonest legs. It is by these means that I have drawn up the tables of measurements of the species. I do not expect that I have escaped error altogether, but I hope that I have arranged my conclusions in such a way that any errors may be readily detected by future investigators. I regret much that I have not been able to make as close an examination of the crania as I could wish, for, with the exception of a few skulls from Shag Point belonging to the genus Euryapteryx, which have been kindly lent me by Mr. Aug. Hamilton, I have only had opportunity for hurried observations in the different museums I have visited. The first announcement of the former existence of large struthious birds in New Zealand was by Mr. J. S. Polack, who, in his book “New Zealand,” published in 1838, stated that he had found their bones near the East Cape of the North Island. Subsequently — 1839 to 1843 — several collections were made by Mr. W. Colenso, the Rev. W. Cotton, and the Rev. W. Williams in that district and at Poverty Bay. In 1844 the Rev. R. Taylor found them in numbers at Waingongoro, near Wanganui, and a large collection was made from here by the Hon. W. Mantell in 1847,* Quar. Jour. Geol. Soc., vol. iv., p. 238; and “Petrifactions and their Teachings,” p. 94. and by Sir G. Grey in 1850. In 1852 Dr. A. S. Thomson found many bones in the limestone caves in the upper Waipa country,† Edin. New Philos. Jour., 1856, vol. lvi, p. 268. and others have been found about Tongariro and Lake Taupo, as well as near Wellington. In 1875 Mr. J. Thorne made a valuable discovery of bones at Pataua, near Whangarei, and others have been since found in that neighbourhood, as well as near Auckland, by Mr. T. F. Cheeseman. In 1887 a rich locality was found at Te Aute, near Napier, and was collected from first by Mr. Aug. Hamilton, and afterwards by Mr. James Park, for the Geological Survey. The first discovery of bones in the South Island was made by Dr. Mackellar and Mr. Percy Earl at Waikouaiti, in Otago, in 1846; a locality which was explored three years later by the Hon. W. Mantell,‡ Quar. Jour. Geol. Soc., vol. vi., pp. 327 and 334; and “Petrifactions and their Teachings,” p. 98. who again, in 1852, made a good collection from an old Maori cooking-place at Awamoa, near Oamaru. In 1861 the limestone caves in Collingwood County were explored by Sir J. von Haast and Mr. Maling, and from this time onwards numerous discoveries were made in Canterbury and Otago. In 1866 Sir J. von Haast began his classical explorations at Glenmark; and the almost equally rich deposit at Hamilton, in Central Otago, was explored by me in

1874. Interesting discoveries have been made lately in caves in the Nelson District, as well as at Castle Rock, in Southland; and only last month another large deposit was found near Oamaru, and collected from by Mr. H. O. Forbes.* For a list of localities at which moa-bones have been found see Geol. Mag., series 3, vol. i., p. 129 (1884). Fragments of egg-shell have been found in abundance at Whangarei and Wanganui,† “Petrifactions and their Teachings,” G. A. Mantell, p. 121. as well as in Canterbury and Otago. The Hon. W. Mantell restored, more or less perfectly, about a dozen eggs from Awamoa;‡ See Owen's Ext. Birds of N.Z., pl. cxv., as crassus. and several have been found in the interior of Otago. A nearly perfect one was found in 1867 at Cromwell, which contained the bones of the embryo chick.§ Trans. N.Z. Inst., vol. iv., p. 110. It measured 9in. by 6in., and was found in sand about 2ft. below the surface. Another, still more perfect, was found at the end of 1859, or the beginning of 1860, at Kaikoura, with a Maori skeleton. It measured 10in. by 7 ¾in., and was sold in London for £120.¶ “Zoologist,” 1866, p. 34; Trans. N.Z. Inst., vol. iv., p. 403; Ext. Birds of N.Z., pl. cxvii. The specimens of shell are almost always white or stained brown with earth; but two fragments of a pale-green colour have been found, one at Queenstown, the other in sand on the banks of the Kawarau River.∥ Trans. N.Z. Inst., vol. viii., pp. 98 and 101; and vol. xviii., p. 83. We may therefore conclude that the egg was probably white in some species, but pale-green, like that of the cassowary, in others. The egg-shell consists of two calcareous layers: the outer layer is the thicker, and is laminated; the inner layer has a columnar structure, each column sometimes containing triangular prisms,** Nathusius, Zeit. Wissensch. Zool., xx., p. 118; and Hutton, Trans. N.Z. Inst., vol. iv., p. 166. but the structure differs much in different eggs. The canals through the shell are oval and simple. Little heaps of stones—moa-stones—from the gizzards of the birds have been often found. These stones vary in size from quite small up to 2oz. in weight, while a complete set weighs from 8oz. up to 51b.†† Ext. Birds of N.Z., pl. xcii.; Trans. N.Z. Inst., vol. xvii., p. 174. Sir J. von Haast noticed that these stones had always been picked up by the birds in the immediate neighbourhood,‡‡ L.c., vol. iv., p. 73. showing that they did not travel fast. Feathers have been found in the alluvial sands of the Clutha River, between Alexandra and Roxburgh, 18ft. below the surface; and also in caves near Queenstown. Most of them have two shafts to each quill, but some have only one

shaft. When there are two shafts to a quill one is usually, but not always, shorter and more slender than the other. The largest are not more than 7in. in length. It would appear probable, from specimens of skin which have been found, that the feathers of the neck were two-shafted, while those on the legs were one-shafted. In colour they differ. Those from the Clutha are brown at the base, shading upwards into black, while the top is white.* Trans. N.Z. Inst., vol. iv., p. 166; Ext. Birds of N.Z., p. 442, and pl. exiv., figs. 8–11. Those from Queenstown are either reddish-brown with a central longitudinal stripe of dark-brown near the apex, or they are pale-brown margined with darker. One was pure-white.† Trans. N.Z. Inst., vol. viii., p. 101, and vol. xviii., p. 83. Evidently the Clutha bird was speckled with white on a black ground, while the Queenstown birds were streaked longitudinally with two shades of brown. A bone of S. casuarinus was found in the cave with the latter feathers, but there is no proof that the feathers and the bone belonged to the same bird. Professor Owen has shown that the legs of M. didinus were feathered down to the toes, the feathers on the metatarsi being from 1in. to 2 ¼in. in length, and without an accessory plume. The basal part of each feather is lightgrey, deepening towards the apex into reddish-brown.‡ Trans. Zool. Soc., vol. ii., p. 257. There were also feathers on the specimen of D. potens found at Tiger Hill.§ Dallas, Proc. Zool. Soc., 1865, p. 265. The feathers are furnished with barbules up to the very tip, except in a few cases; but barbicels are always absent. The shafts are slender and flexible, and do not project beyond the barbs. Not only were the moas at one time very numerous, but there were many different kinds, ranging in height from 2 ½ft. up to 11ft. or 12ft.—that is, 3ft. or 4ft. higher than the largest ostrich; but the smaller birds were far more common than the large ones. They all agreed in being remarkably robust in build, with strong legs, and rather flat heads with small eyes but well-developed olfactory organs. The whole bird, from the head downwards, was covered with soft fluffy feathers. There were no long plumes on the head or on the tail. Dr. von Hochstetter, in his book “New Zealand,” has given, on page 176, an admirable restoration of one of the species of Dinornis. The brain was very small in proportion to the size of the spinal cord.∥ Ext. Birds of N.Z., p. 326, pl. xci., figs. 11–13. The beak differed a good deal in the different kinds, being sharp and pointed in some, obtuse and rounded

in others, but it was never much longer than the head. The neck was remarkably long, having nearly as many vertebræ as a swan, and was very strong and muscular. The body was short and plump, without any wings; while the legs were very massive, and with particularly muscular thighs. The length of the leg differed very much; and some of the South Island species must have been absurd-looking creatures—nearly as broad as high, with short, heavy, stumpy legs, on which they could hardly waddle. Probably all of them had a hind-toe which was raised from the ground. The claws were strong, and formed good implements for scratching. Those of the front toes were curved, while that of the hind-toe was nearly straight. Like other struthious birds, they voided their urine separately, and dry fæces have been found in a cave near Queenstown, which proved, what had previously been inferred from the gizzard-stones, that the moas were vegetable feeders.* Taylor White, Trans. N.Z. Inst., vol. viii., p. 99. Evidently the birds were stupid and sluggish, and probably diggers of the ground. Very little is known about the soft parts, but the integument, tendons, and some of the muscles of the leg and foot of D. potens,† Ext. Birds of N.Z., p. 248, pl. lxxi.; Trans. N.Z. Inst., vol. vii., pp. 267 and 271. S. crassus,‡ Trans. N.Z. Inst., vol. vii., p. 143. and M. didinus§ Trans. Zool. Soc., vol. xi., p. 257. have been described by Sir R. Owen, by Dr. Coughtrey, and by myself. The integument and muscles of the neck of S. crassus∥ Trans. N.Z. Inst., vol. iv., p. 114; and vol. vii., p. 141. have been described by Sir James Hector and by Dr. Coughtrey; and the head of M. didinus¶ Trans. Zool. Soc., vol. ii., p. 257. by Sir R. Owen. Two fragments of egg-shell found in Monck's Cave, near Sumner, have the shell-membrane still remaining.** The “internal epidermis” of Mr. Forbes. Trans. N.Z. Inst., vol. xxiii., p. 374. It is very thick, fibrous, and stained brown: the egg-shell is white. Classification of the Moas.†† I regret that I have not been able to see Mr. Lydekker's “Catalogue of the Fossil Birds in the British Museum,” 1891. No copy has as yet been received in Christchurch. Sir Julius von Haast, in 1874, divided the moas into two families—the Dinornithidæ and the Palapterygidæ, characterized by the absence or presence of a hind-toe, and by a supposed difference in the osseous tissue of the bones; those of the Palapterygidæ being, he thought, rougher on the exterior, harder, thicker, more dense, and consequently heavier, than

those of the Dinornithidæ.* Trans. N.Z. Inst., vol. vi., p. 424. In Nicholson's “Manual of Palæontology,” vol. ii., p. 1226, Mr. Lydekker says, “According to the late Sir J. von Haast, the Palapterygidæ were provided with rudimentary wings.” This is a mistake, for Dr. Haast distinctly says that his Palapterygidæ had not even a scapulo-coracoid. In this he has been followed by Dr. H. Woodward, † Proc. Geol. Association, 1885, p. 363. Dr. Woodward's reference in this paper to Erythromachus as belonging to the Apterygidæ is a mistake. It is a carinate ralline bird, thought by A. Milne-Edwards to be allied to Ocydromus. and doubtfully by Professor A. Newton, who included the two families in a separate order, which he called “Immanes.”‡ Encyclopædia Britannica, 9th ed., art. “Birds,” p. 73, and art. “Ornithology,” p. 44. But long ago I pointed out that these characters would have been too slight to distinguish the families even if they had been correct;§ Trans. N.Z. Inst., vol. ix., p. 363. and it is now known that most, probably all, of the moas had hind-toes, and that no distinction in osseous tissue can be made out. I therefore include them all in one family. Genera. Sir R. Owen founded the genus Dinornis in 1843, on the leg-bones of D. struthioides¶ Proc. Zool. Soc., 1843, p. 8; and Trans. Zool, Soc., vol. iii., p. 235. and several other species. In 1846 he formed a new genus—Palapteryx—to receive those species in which he had inferred the existence of a hind-toe—namely, P. ingens, and P. dromioides. In 1848 he described a cranium as that of P. geranoides (really P. dromioides), and another as P. dromioides (probably C. geranoides). Still later he unfortunately described the skull of Aptornis as that of Dinornis casuarinus, and he thus made a wide difference between Dinornis and Palapteryx. The mistake was pointed out by Professor W. K. Parker; and Owen abandoned his genus Palapteryx, not considering that the absence or presence of a hind-toe was, by itself, sufficient to distinguish the two genera.¶ Ext. Birds of N.Z., p. 258. In 1850 Reichenbach, in his “Das Natürliche System der Vögel,” made six new genera out of the eight species which Professor Owen had described up to that time, so that each genus contained one species only. I have not seen this book, and know it only by Owen's quotation,** Ext. Birds of N.Z., p. 417. but it would seem that Reichenbach shot at a venture, because he could not possibly have given good characters to his genera. In 1874 Sir J. von Haast grouped the moas of the South Island into four genera—two in his Dinornithidæ and two in his

Palapterygidæ—and gave several generic characters by which they might be distinguished, some of which were good, but most of them quite incorrect. In fact, at that time the species had not been accurately made out,* See, however, Trans. N.Z. Inst., vol. iv., p. 70, line 9 from top. and consequently a correct grouping of them was impossible. Sir R. Owen subsequently reviewed the whole question, and came to the conclusion that it was best for the present to retain all the species in the single genus Dinornis, and this he has done in his “Extinct Birds of New Zealand,” 1879. But undoubtedly the differences in the crania of the different moas are quite sufficient to indicate the existence of several genera; the difficulty hitherto in defining these genera has been that the cranial differences did not appear to go with the other characters. But the extensive material examined by me has shown that in many cases the heads have been placed on the wrong birds, the exceptions being D. robustus, D. struthioides, D. parvus, D. didinus, and D. gravis; and after reducing the species to order I find that they fall into seven well-defined genera, founded on the crania, but generally accompanied by characters derived from the pelvis, the sternum, the absence or presence of a scapulo-coracoid, and the robustness of the leg-bones. Species. Nineteen species have already been made by Sir R. Owen, and one by Sir Julius von Haast, the specific characters being based on the bones of the leg. However, only four of these species—viz., D. maximus, D. didinus, D. parvus, and D. oweni—have been described from the bones of a single individual. In all other cases, the different bones of the leg had to be put together conjecturally from collections sent to England at different times between 1843 and 1872, and so it is no wonder that many of the bones were wrongly placed. For example, the examination of the legs of individual birds has shown that D. altus, D. ingens, D. gracilis, D. struthioides, D. geranoides, D. rheides, D. crassus, D. casuarinus, and D. gravis, of Owen, are all made up of bones belonging to more than one species. In these cases I have taken the metatarsus as the type of the species, except with D. ingens, where the tibia forms the type. These individual legs, together with a comparison of a very large number of bones from all parts of New Zealand, has enabled me, I believe, to correct most of the errors, and to reduce the species to order. The most difficult bones are the femora. Indeed, with our present limited knowledge it is sometimes impossible to refer them with confidence to the right species.

The species are not distinguished by any anatomical differences in the leg-bones, but merely by size. In many cases, especially in the genera Syornis and Euryapteryx, there are considerable differences in the amount of divergence of the metatarsal condyles, which may be of specific value, but I am inclined to think that it depends largely on the age of the individual. Also, the tibiæ vary much in the curvature of the shaft; but there are complete gradations from one form into the other, and I cannot think this curvature to be of specific value.* Trans. N.Z. Inst., vol. vii., p. 277. There appear to be specific differences other than size in the crania; but this must be left for future investigation. So far as the leg-bones are concerned, my examination has shown that several of the species pass gradually one into another, so that any line separating them must be an arbitrary one; and I should not be surprised if further knowledge showed that this applies to those species which at present appear to be distinctly marked off. For we are here dealing with a large number of individuals which lived for a very long time without the check of natural selection to eliminate the intermediate forms—a case which is probably unique in natural history. In the tables of measurements of each species I have given first the measurement of Professor Owen's type, then those of individual leg-bones found together, and then the maximum and minimum size and robustness which I allow to each species—the latter being represented by the girth at the middle of the shaft. It may be thought that the twenty-six species which I have admitted are too many; but after a careful consideration of the subject I have come to the conclusion that a less number would not represent the facts sufficiently, and would not give an adequate idea of the great variety of form which the moas exhibit. In the genus Syornis only the smaller individuals are found in the North Island, and this gives us a maximum limit for S. casuarinus. The larger individuals of the genus, which are found only in the South Island, show too great a diversity to be included in one species at all uniform with S. casuarinus, and consequently they must be divided into two. From this we get an idea of the amount of variation to be allowed to each species in the genera Euryapteryx and Cela. With Dinornis and Palapteryx the case is different, because with them we have to take geographical distribution into consideration, and smaller morphological distinctions may with advantage be taken to differentiate species when they are combined with a different geographical distribution. It is also a question which must be left for future investigation whether those

species in the two Islands which are nearly of the same size are truly representatives of each other, or whether they form two parallel lines of development, one in each Island, both of which are descended from D. struthioides. To join the species together would delay the settlement of this question. Sexual Differences. In his second paper on Dinornis, Professor Owen discussed the question whether some of his species might not be the opposite sex of others, and came to the conclusion that they were not so.* Ext. Birds of N.Z., p. 90. With this conclusion I quite agree. A. didiformis has no near ally, and must therefore contain individuals of both sexes. It is the same with Palapteryx, of which there is only one species in each Island. S. casuarinus and M. didinus are not likely to be male and female, because their heads differ so much; and, if they are not a pair, each of these species must contain individuals of both sexes in the North Island, and, consequently, in the South Island also. If D. struthioides is only one sex, then the opposite sex must have had different forms in the two Islands—D. torosus in the South, and D. gracilis in the North. But D. torosus and D. struthioides cannot be the two sexes of one species, because they have different geographical distributions, D. torosus being almost entirely confined to the north part of the South Island, while D. struthioides was more abundant in the south than in the north. It is the same with C. geranoides and C. curtus, the former being more abundant in the south-eastern, the latter in the north-western, part of the North Island. And if this is true for six of the genera we may be sure that it is true for Euryapteryx also. Sir Julius von Haast has on several occasions stated his opinion that several species of moa consist of individuals of two distinct sizes, with no intermediate forms, which he took to be male and female, and he refers to specimens in the Canterbury Museum as confirmatory of this opinion.† Trans. N.Z. Inst., vol. i., p. 81, and vol. vi., p. 425; Geol. Cant. and West., 1879, p. 433. Unfortunately, he never published any measurements showing these supposed sexual differences, and the collection he refers to in the Canterbury Museum by no means bears out his statement. Dr. Haast had specially named S. casuarinus as one that showed these sizes, and I therefore published, in 1874, measurements of all the bones of this species found in the swamp at Hamilton;‡ Trans. N.Z. Inst., vol. vii., p. 278. and these show a gradual passage from the smallest to the largest. In the following year Mr. Thorne

published measurements of all the bones found by him near Whangarei,* Trans. N.Z. Inst., vol. viii., p. 94. which proved the same thing for C. curtus. Nevertheless, Sir Julius repeated his statement in 1879 without referring to either of these lists of measurements; and in 1886, in his paper on D. oweni, he oncemore says that there are two different sizes in that species.† Trans. Zool. Soc., vol. xii., p. 172. But again he gives no measurements in proof of his statement; and the extensive table which I have compiled of the measurements of C. curtus and C. oweni, as well as of several other species, shows positively that they are not divided into two distinct sizes differing either in the length or in the robustness of their limbs. Certainly I admit two varieties, differing in size, in S. casuarinus; but the sizes overlap one another, and have been recognised only because they have a different geographical distribution; which is, by itself, a sufficient proof that they cannot be the sexes of one species. I do not, of course, maintain that there was no difference in size between the sexes. In the ostrich, the emu, and the cassowary the male is rather larger than the female; but in the kiwi the female is considerably larger than the male. Professor T. J. Parker has shown that in the kiwi the hind limbs undergo a relative diminution in size between the time of hatching and the attainment of fully adult proportions,‡ Phil. Trans., vol. clxxxii., p. 42, and pl. vi. and this is especially the case with the female. This implies that the ancestral kiwis were, like Megalapteryx, larger than the living birds; and we may infer the same thing from the great size of the egg. It is a legacy from a larger bird, which is not easy to get rid of. The greater proportionate size of the female is therefore probably due to its having to lay such a very large egg. The males have decreased in size more rapidly than the females, who were handicapped by such large eggs. I shall subsequently show that the moas increased in size during development, so that the usual rule of the male being larger than the female would probably apply to them. But this difference has not yet been proved; and, if it existed, some individuals of the larger sex must have been as small as, or smaller than, some individuals of the smaller sex. Family Dinornithidæ. The skull is remarkable for the great breadth and flatness of the occipital region, and the broad and deep temporal fossæ. The plane of the foramen magnum is nearly vertical, and the basi-occipital forms nearly a right angle with the basi-sphenoid. There is a broad and deep depression between the condyle and

the paroccipital process. In the interior of the brain-case the tentorial ridge is low, as also is the bony plate between the cerebral and optic fossæ. The optic foramina are widely separated. The olfactory cavities are very large, extending backward between the orbits nearly to the optic foramina, each with a separate olfactory foramen. The lateral ethmoids are large, but do not appear on the upper surface of the skull; they are rolled up into a small spiral process in the posterior portion of each olfactory chamber, forming the posterior turbinals. The internasal septum is perforated posteriorly. The pterygoids articulate with basi-pterygoid processes, which arise from the basi-sphenoid. The rostrum is very long; the ventral border of the presphenoid is produced on each side into a horizontal triangular plate. The vomers are broad and separate; they articulate posteriorly with the pterygoids, and are connected with the palatines. The palatines join the vomers posteriorly, and also articulate with the pterygoids; the maxillary processes are produced forward, and end anteriorly in horizontal expansions, which are interposed between the maxillaries and the vomers; there are no distinct maxillo-palatine processes. The upper squamosal head of the quadrate is broad and single. The nasals anchylose with each other posteriorly. The beak is rather short, never much longer than the head, more or less curved downwards. The eye had a ring of bony sclerotic plates. The vertebral column consists of from 54 to 56 vertebræ, as follows: Cervical, 20 or 21;* Hutton, Ann. and Mag. Nat. Hist., ser. 5, vol. i., p. 407, and vol. ii., p. 494 (1878). Owen, Trans. Zool. Soc., vol. xi., p. 235 (1882). Haast, Trans. Zool. Soc., vol. xii., p. 174 (1886). In Owen's Ext. Birds of N.Z., pp. 392, &c., the 3rd cervical is really the 4th, the 4th is the 6th, the 6th is the 8th, the 12th is the 15th, the 14th is the 18th, and the 15th is the 21st. cervico-thoracic, 3 or 2; thoracic, 7, of which the last three or four are united to the pelvis; lumbar and sacral, 10 or 11; caudal, 14 or 15, of which the first three or four are united to the pelvis, and the last two or three are anchylosed into a single bone which gradually tapers to an obtuse point and is of less vertical height than the preceding free caudals. The syn-sacrum (of Professor T. J. Parker) consists of seventeen or eighteen vertebræ. In the cervical vertebræ the median inferior process, which commences on the axis, becomes obsolete on the sixth, and is never present on the seventh. Hypapophyses commence on the seventh, the two laminæ are furthest apart on the fifteenth or sixteenth, then approach again and fuse on the eighteenth or nineteenth without forming a hæmal canal. The neural spine, which is single on the axis, divides into two laminæ on the third, which usually remain separate until the

nineteenth or twentieth, when they coalesce. The dorsal vertebræ are remarkable for the thick neural spine and the small neural canal. The centrum is saddle-shaped in all of them. There are nine or ten pairs of vertebral ribs, of which the last three or four pairs are attached to the pelvis. The first two or three pairs are cervico-thoracic, are often very short, and have no sternal ribs. The thoracic ribs have sternal ribs, of which the first and second pairs, and in most genera the third pair also, are connected with the sternum. The true ribs, and some of the floating ribs, bear uncinates, which are often anchylosed to the rib. These uncinates are of nearly equal breadth throughout, and are not expanded at the proximal end: their length is from 1.5 to 2.5 times the breadth. The sternum is broad and flat, with its anterior border slightly concave. The costal (anterior xiphoid) processes are nearly horizontal. There is one long lateral (posterior xiphoid) process on each side, and the posterior median process is often notched at the end. Coracoid grooves are reduced to shallow oval pits, and may be absent altogether, or developed on one side only. It is a very variable bone, and often unsymmetrical. There is no median centre of ossification. A scapulo-coracoid is present in three genera and absent in the others. The coracoidal portion is subcylindrical, rounded at the proximal end, and attached to the sternum by ligament only. The scapular portion is nearly one and a half times the length of the coracoid: it is slightly flattened and curved in the plane of the coraco-scapular angle only. This angle varies between 145° and 125°, and the bone is sometimes flattened and broadened at the angle. There were no clavicles. Wings appear to have been present in Palapteryx only. The pelvis is broader than that of any other ratite bird. The ilia commence to diverge a little in front of the acetabula, and come together again, enclosing a broad oval sacrum. The ischia and pubes are generally free for their whole length, but sometimes they are attached together for a short distance a little before the middle, leaving a small obturator fissure; and occasionally (? in very old birds) the posterior ends are united, but the ischia always remain free from the ilia. There is no pectineal process. The femur is nearly straight anteriorly, and is distinguished by the tuberosities on the posterior part of the shaft, by the trochanter rising above the articular head, by the great size of the distal end and the great breadth of the rotular surface. There is no pneumatic foramen, and the dense bony wall of the shaft is very thick.* Ext. Birds of N.Z., pl. xxix., fig. 1. The patella† Trans. Zool. Soc., vol. ii., pl. li. and lviii. is bony and trihedral;

the outer surface broadest, curved, and rather rough, the two inner surfaces smooth. The tibia (tibio-tarsus) is distinguished by the breadth of the cnemial process, by the oblique bony supratendival bridge at the inner anterior surface at the distal end, by the smallness of the epicondylar tuberosities, and by the prominence anteriorly of the inner condyle. The fibula is not anchylosed to the tibia. There is a free post-axial tarsal bone—the second centrale* Parker, Phil. Trans., vol. clxxxii., p. 100, footnote; Ext. Birds of N.Z., pl. cxiv., figs. 3–6 (sesamoid). Coughtrey, Trans. N.Z. Inst., vol. vii., p. 269 (calcaneo-sesamoid). Buller, Birds of N.Z., 2nd ed., vol. i., p. xxxii., and vol. ii., p. 333 (astragalus).—on the posterior and outer side of the tarsal joint. The metatarsus (tarso-metatarsus) is characterized by the smooth and rounded surfaces of the shaft. At the proximal end the depression for the articulation of the inner condyle of the tibia is deeper than that for the outer one. The hypotarsal process is grooved, and the outer side of the groove is more prominent than the inner. The interosseous canals from the two posterior foramina unite and open anteriorly by a single foramen. There are no perforations at the distal end of the metatarsal grooves.† In one specimen of D. firmus I have seen an osseous bridge over the channel between the middle and inner metatarsal condyles, the perforation answering to that between the second and third metatarsals in Megalapteryx. The inner condyle is more produced than the outer one. A hind-toe was probably present in all the species; and there was a sesamoid on the lower surface of the foot, between the metatarsus and the first phalanx of the inner toe.‡ Ext. Birds of N.Z., pl. lxxi., fig. 1s. The inner toe has three, the middle one four, and the outer toe five phalanges. Measurements. In the table of ratios in the genera, those for the leg-bones are the length (L) divided by the girth (G) at the middle of the shaft. For the skull, the breadth at the squamosals (B) is divided by the height (H) at the basi-temporal, and the length from the supra-occipital to the nasals (L) is divided by the breadth at the squamosals. For the sternum, the length of the body (L), measured from the anterior border to a point midway between the two posterior notches, is divided by the breadth of the body (B) just below the costal region. For the pelvis, the length of the ilium (L) is divided by the length of the pre-acetabular portion (l), and the breadth at the antitrochanters (B) is divided by the length (l) of the pre-acetabular portion of the ilium.

Table of Ratios in the Genera. Metatarse. Tibia. Femur. Skull. Sternum. Pelvis. Coracoid. L:G L:G B:H L:B L:B L:l B:l Dinornis Max. 3.6 5.7 2.5 2.1 0.9 0.6 2.3 1.0 * Min. 2.5 4.3 1.7 2.0 0.8 1.9 0.9 Tylopteryx Max. 3.3 5.5 2.2 2.4 0.8 0.6 2.2 1.1 * Min. 2.7 4.6 1.9 1.8 0.7 2.0 0.9 Palapteryx Max. 3.2 5.4 2.9 1.8 1.1 0.7 1.8 0.7 * Min. 2.5 4.6 2.4 Anomalopteryx Max. 2.3 5.1 2.5 1.6 1.1 1.0 2.0 0.9 * Min. 2.0 4.2 2.0 1.5 1.8 Cela.. Max. 2.2 4.4 2.4 1.7 1.1 0.9 2.5 1.2 0 Min. 1.7 3.8 1.8 1.6 2.4 1.1 Mesopteryx.. Max. 2.0 4.0 2.2 1.5 1.1 0.7 2.3 1.2 0 Min. 1.7 3.7 1.8 Syornis Max. 2.2 4.6 2.1 1.7 1.2 1.1 2.5 1.3 0 Min. 1.6 3.7 1.9 1.6 1.0 0.7 2.3 1.2 Euryapteryx.. Max. 1.5 3.7 1.8 1.6 1.2 0.6 2.7 1.5 0 Min. 1.3 3.0 1.5 1.5 1.1 0.5 2.4

In the tables of measurements of the species, the girth in the leg-bones is taken in the middle of the shaft. The length of the pelvis is that of the pre-acetabular portion of the ilium only, and the breadth is taken at the anti-trochanters. The breadth of the sternum is taken across the body just below the costal region. The length of the skull is taken from the supra-occipital protuberance to the commencement of the nasals, the depression for which is often preserved; the breadth of the skull is taken at the squamosals, and the height is the vertical from the basi-temporal. Other measurements are given when known, and explain themselves. In the tables of measurements given for each species, an exclamation point (!) prefixed means that all the measurements along that line are taken from the same individual bird. If no exclamation point is prefixed the measurements given are only inferred to belong to bones of the same species. The maximum and minimum measurements give the limits in size, and in robustness of the legs, allowed to each species. None of the measurements given are hypothetical. All are either from actual specimens or from Professor Owen's plates, with the single exception of the length of the metatarsus in A. antiquus. Genus Dinornis.* Greek δεlυóς, terrible. Dinornis (part) and Palapteryx (part) of Owen; Dinornis, Movia, and Moa of Reichenbach; Dinornis of Haast. Skull depressed, the lambdoidal ridge flattened, and the parietals hardly rising above it. Breadth at the squamosals about twice the height from the basi-temporal. Length from the supra-occipital to the nasals less than the breadth at the squamosals. Occipital condyle pedunculate, hidden laterally by the paroccipital processes. Beak rather longer than the head, depressed and obtuse at the tip; the lower jaw much curved. Sternum with well-marked oval coracoid pits; the length of the body† The “corpus-sterni” of Prof. T. J. Parker, Phil. Trans., vol. clxxxii., p. 85. rather more than one-half the breadth; costal processes slightly developed; the lateral processes long, diverging from the middle line at an angle of between 35° and 50°, but curving inwards again;‡ See, however, the remark under D. firmus. median process generally notched at the end. Only two pairs of sternal ribs articulating with the sternum. Scapulo-coracoid always present, without any glenoid cavity.

Pelvis comparatively narrow, the greatest breadth at the anti-trochanters being equal to or slightly less than the pre-acetabular portion of the ilium. The whole length of the ilium is from twice to two and a quarter times the length of the pre-acetabular portion. Metatarsus longer than the femur, its length between 2.5 and 3.6 times the girth at the middle of the shaft. Tibia rather less than twice the length of the metatarsus, its length between 4.3 and 5.7 times the girth. The femur is about 0.9 times the length of the metatarsus, its length between 1.7 and 2.5 times the girth. The presence of a hind-toe has been proved in D. robustus, D. potens, and D. torosus. Sub-Genus Dinornis. Top of the skull flattened. Temporal fossæ very broad, about equal to the orbits. Breadth of the skull at the squamosals from 1.1 to 1.2 times the length to the nasals. Length of the basi-sphenoid from the basi-occipital to the end of the rostrum about two and a quarter times the breadth at the basipterygoid processes. Occipital condyle exposed from above. Length of the premaxillæ to the maxillary suture about equal to the breadth. Plate XV., fig. 1. N.B.—The North Island species are admitted into this sub-genus provisionally, as their skulls are not yet known. Dinornis altus. D. maximus, Owen, Ext. Birds of N.Z., p. 253 (Dr. Lillie's specimen), metatarsus only. D. altus, Owen, Ext. Birds of N.Z., p. 361 (1879). Figures.—Metatarsus, Ext. Birds of N.Z., pl. lxxix., fig. 4, dotted outline only. Metatarsus: Length, 21.5in.; girth, 6.3in. This species is as yet only known by a single metatarsus from the South Island, the exact locality being unknown. As it is not only longer but also more slender than the similar bone in D. maximus, I agree with Professor Owen in keeping it separate. The femur associated with it by Owen belongs to D. validus. Dinornis maximus. D. maximus, Owen, Trans. Zool. Soc., vol. vi., p. 497 (1867), and vol. x., p. 147; Ext. Birds of N.Z., pp. 250 and 391. D. maximus, Haast, Trans. N.Z. Inst., vol. i., p. 87, No. 18. Figures.—Metatarsus, Ext. Birds of N.Z., pl. lxxix., fig. 3; tibia, l.c., pl. lxxx.; femur, l.c., pl. lxxix., fig. 1. Restoration, l.c., pl. xcvii.

Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type! 20.0 8.3(?) 39.0 8.5 18.25 9.5 No.1! 39.2 8.7 18.4 9.6 Max. 20.5 6.75 39.2 8.7 18.75 9.6 Min. 20.0 6.4 39.0 8.0 18.0 9.2 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type! No. 1! Max. Min. Length of the scapulo-coracoid, from 10.25in. to 9.5in. Distribution.—This very rare species has been found only at Glenmark, with the exception of a single femur, 18in. in length, collected by Mr. H. Joseph in Otago, and taken to England. The type was a metatarsus,*There appears to be a mistake in the measurements given of this bone. obtained in August, 1865, by Major Michael, at Glenmark, and subsequently the femur and tibia of, apparently, the same bird were found in the same locality. No. 1 was obtained at Glenmark by Sir J. von Haast. This is not the original D. giganteus, var. maximus, of Owen, but a larger species, to which he afterwards transferred the name. Dinornis excelsus. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Max. 20.0 6.25 38.0 Min. 20.0 5.5 37.5 7.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Max. Min.

This species is known only by three metatarsi collected by Mr. A. Hamilton in the Te Aute Swamp, near Napier, and by a tibia from the same locality, the dimensions of which have been kindly sent to me by Mr. Colenso. Mr. Hamilton also informs me that two or three tibiæ, with a length of 38in., were obtained by other collectors. It is a smaller and more slender species than D. maximus, but considerably larger than D. giganteus. Dinornis validus. D. maximus, Owen, Trans. Zool. Soc., vol. x., p. 147 (part). D. giganteus, Haast, Trans. N.Z. Inst., vol. i., p. 88, No. 20. Figures.—Femur, Ext. Birds of N.Z., pl. xxxvi. and xxxvii., fig. 1 (as giganteus); vertebræ described and figured on p. 391, &c. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 18.9 6.8 35.0 7.0 16.5 8.7 Max. 19.0 6.8 35.5 7.7 17.0 8.7 Min. 18.0 6.0 34.0 6.5 16.0 7.5 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. 4.0 5.0 2.2 Min. 3.7 4.5 2.2 Breadth of skull at temporal fossæ, 2.8in.; at post-frontal processes, 4.6in. Length of the lower jaw, 7.85in. Scapulocoracoid, 8.7in. to 8in. Middle toe, 11in. Distribution. — In the swamps at Glenmark and Waikouaiti, and in a few other places in Canterbury, Otago, and Southland. It seems to have been more common in the north part of the South Island, and rare in the south. No bones were found at Hamilton. The type is from Glenmark. This species is larger and rather more robust than D. giganteus. Earl's specimens from Waikouaiti—the original types of D. maximus—belong here. I have already mentioned that Owen transferred the name to a larger species, and I think it better to follow him in the change, as it will make less confusion.

Dinornis giganteus. D. giganteus, Owen, Trans. Zool. Soc., vol. iii., p. 237 (1843), and.p. 307. (1846), and vol. iv., p. 59 (1850). Moa giganteus, Reichenbach, Das Nat. Syst. der Vögel. Figures.—Metatarsus, Ext. Birds of N.Z., pl. xxvii., fig. 1. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 18.5 5.5 35.0 6.5 16.0 7.25 No. 1! 17.75 6.75 31.5 7.25 Max. 18.5 6.75 36.0 7.25 16.0 7.5 Min. 17.5 5.0 34.2 6.5 16.0 7.25 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! Max. Min. Length of the scapulo-coracoid, 8–75in. Distribution.—The type of this rare species is from Poverty Bay. Mr. A. Hamilton obtained a few bones from the Te Aute Swamp. No. 1 was found in a swamp at Awhitu, near Auckland, by Mr. Mactier.* Trans. N.Z. Inst., vol. x., p. 552. The only bones found were a pair of metatarsi, a pair of tibiae, and a number of toe-bones The cnemial crest of the tibia is probably abraded. The bones are now in the Auckland Museum. The metatarsi in this species and in D. excelsus are more slender than in any other of the moas. Dinornis robustus. Palapteryx robustus, Owen, Trans. Zool. Soc., vol. iii., p. 345 (1848); vol. iv., pp. 1 and 337 (cranium from Timaru); Mantell, Quar. Jour. Geol. Soc., vol. vi., p. 338. Figures.—Metatarsus and foot, Ext. Birds of N.Z., pl. xlix.; cranium, pl. lxii. and lxiii.; sternum (?), pl. xxxv., fig. 1.

Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 15.7 5.2 32.2 6.7 No. 1! 16.2 6.3 32.7 6.2 15.5 7.9 No. 2! 17.2 6.0 31.5 6.2 14.5 8.1 Max. 17.2 6.7 32.7 6.8 15.5 8.1 Min. 15.7 5.2 30.0 6.2 14.4 7.2 Pelvis. Sternum. Skull Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 10.0 10.0 8.5 No. 2! 10.75 10.0 8.3 Max. 10.75 10.0 8.6 3.85 4.25 2.1 Min. 9.5 10.0 8.0 3.6 4.10 2.0 Breadth of the skull at temporal fossae, 2.9in. to 2.7in.; at post-frontal processes, 4.6in. Length of the lower jaw, 7.6in. Length of scapulo-coracoid, 7.25in. Length of ilium 22in. to 21.5in., of ischium 12in. to 10.2in.; breadth of sacrum, 8in. to 7.5in.; depth of pelvis, 9in. to 8in. Post-axial, tarsal is 2.6in. in length, and 1.5in. in greatest breadth. Length of middle toe, 11.5in. to 9.75in. Total length of head, 8.5in. Distribution.—Found sparingly all over the South Island, but rather more abundantly in Otago than elsewhere. There is a femur from near Greymouth in the Wellington Museum. Bones are found not only in the swamps but occasionally in the Maori cooking-places. No. 1 is from Shag Valley, and No. 2 from Highley Hill, in Otago: both are in the Otago Museum. The type is from Waikouaiti. Sir R. Owen originally gave the name D. ingens, var. ro-bustus, to bones from Waikouaiti, of which the metatarsus was 14–5in. and the tibia 28.75in. in length; but he afterwards transferred the name to larger bones from the same locality, and said that the tibia of his type of D. robustus is 32in. in length.*Ext. Birds of N.Z., p. 250. Under these circumstances it seems best to retain the name for this species, especially as it is so used in New Zealand. In specimen No. 1 the neural spine is single on the fourth and fifth cervical vertebræ.

Dinornis firmus. D. ingens, var. robustus, Owen, Ext. Birds of N.Z., p. 160, from Wanganui. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. No. 1! 16.25 5.6 29.0 6.25 14.5 7.5 Max. 17.0 5.6 33.0 6.5 15.25 7.5 Min. 15.75 4.75 30.0 5.3 14.5 6.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. No. 1! Max. 8.25 Min. Distribution.—Rather common; the swamp at Te Aute, also at Poverty Bay, and in the Maori cooking-places at Wanganui and Whangarei. No. 1 is a specimen in the possession of W. Colenso, Esq., from Poverty Bay. The cnemial crest of the tibia is probably abraded. This was a more slender species than D. robustus, but occasionally bones are found which cannot be distinguished from slender South Island specimens. A fragmentary sternum in the Napier Museum has the lateral processes diverging at a less angle than that of robustus, which may also indicate a specific difference. The tibiæ of this species and of D. ingens are more slender than those of the other moas. Dinornis ingens. D. ingens, Owen, Trans. Zool. Soc., vol. iii., p. 237 (1843); Movia ingens, Reichenbach, Das Natürliche System der Vögel. Figures.—Tibia, Ext. Birds of N.Z., pl. xxv., fig. 1 and pl. xxxvi., figs. 1 and 2; femur, l.c., pl. xxi., figs. 1 and 2; skull, l.c., pl. liii., figs. 1–3, juv. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 29.0 5.25 13.0 5.5 Max. 15.25 4.75 29.5 5.3 14.25 7.25 Min. 14.6 4.3 27.0 5.0 13.0 5.5

Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. Min. Length of the scapulo-coracoid, 7in. Distribution.—At present this species is only known from the Te Aute Swamp and Poverty Bay, from whence the type specimens came. But the skull, from the Taupo district, figured on plate liii., figs. 1–3, of the “Extinct Birds of New Zealand,” probably belongs to a young individual. The metatarsus assigned to this species by Professor Owen belongs to D. gracilis. Dinornis potens. D. ingens, var. robustus, Owen, Trans. Zool. Soc., vol. iii., p. 307 (1846); Ext. Birds of N.Z., p. 129. Pal. robustus, Owen, Trans. Zool. Soc., vol. iv., p. 337, skull and scapulo-coracoid from Manuherikia; vol. vi., p. 495, foot. D. ingens, Haast, Trans. N.Z. Inst., vol. i., p. 84, No. 8; Hutton, l.c., vol. vii., p. 278; Hutton and Coughtrey, l.c., vol. vii., p. 266, pl. xix. Figures.—Skull and scapulo-coracoid, Ext. Birds of N.Z., pl. lxiv., lxv.; integument of foot, l.c., pl. lxxi.; Trans. N.Z. Inst., vol. vii., pl. xix. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. No. 1! 15.1 5.7 29.5 6.0 14.2 7.6 No. 2! 15.0 5.4 29.7 6.5 13.0 7.75 No. 3! 15.0 27.7 5.75 13.75 7.0 No. 4! 15.0 5.5 27.0 6.0 14.0 7.0 Max. 15.1 5.7 29.5 6.5 14.2 7.75 Min. 14.0 5.0 27.0 5.6 13.0 7.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height No. 1! No. 2! No. 3! No. 4! 7.5 3.5 4.1 2.1 Max. 8.5 9.5 3.6 4.1 2.15 Min. 3.5 4.0 2.1

Breadth of skull at temporal fossae, 2.8in. to 2.6in.; at post-frontal processes, 5.15in. to 4.9in. Length of lower jaw, 7in. to 6.8in. Total length of the head, from 8in. to 7.8in. Length of scapulo-coracoid, 7.5in. Length of ilium, 20in. to 19.5in.; breadth of sacrum, 7.75in.; depth of pelvis, 7in. Distribution.—Found all down the east side of the South Island from Motunau to Southland, but most common in Otago. No. 1 is from Heathcote, near Christchurch; No. 2 from a cave at Castle Rock, Southland (it is now in the possession of Mr. Mitchell); No. 3 is the leg from the Knobby Ranges, Otago, with integument attached;*Trans. N.Z. Inst., vol. vii., p. 266. No. 4 from Tiger Hill; Manuherikia, Otago. For the measurements of this last specimen I am indebted to Mr. H. M. Platnauer, curator of the York Museum. I am not aware of any bones having been obtained from Maori cooking-places. This species is more robust than D. ingens and smaller than D. robustus. Sub-Genus Tylopteryx.† Greek, ύλος, a swelling, in allusion to the top of the skull. Top of the skull more or less elevated. Breadth of the temporal fossae about two-thirds that of the orbits. Breadth of the skull at the squamosals about 1.3 times the length to the nasals. Length of the basi-sphenoid, from the basi-occipital to the end of the rostrum, about two and three-quarter times the breadth at the basi-pterygoid processes. Occipital condyle hidden by the supra-occipital. Beak rather narrower and not so obtuse at the tip as in Dinornis; length of the premaxillse to the maxillary suture about one and a quarter times the breadth. Plate XV., fig. 2. Dinornis gracilis. D. gracilis, Owen, Trans. Zool. Soc., vol.. iv, p. 141 (1854), metatarsus only. Figures.—Cranium, Ext. Birds of N.Z., pl. lii. (as ingens); metatarsus, l.c., pl. xl., fig. 1 (as ingens), and pl. liv., fig. 3. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 13.0 4.25 Max. 14.25 4.9 26.5 5.5 12.75 6.75 Min. 13.0 4.25 25.0 4.6 12.0 5.75

Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. 3.2 3.7 1.8 Min. Breadth of the skull at the temporal fossæ, 2.65in. Length of the scapulo-coracoid, 6.25in. to 5.7in. Total length of the head, 7.3in. Distribution.—The type of the species is from the sand-dunes near Wanganui. It was plentiful in the Te Aute Swamp, and has been also found at Poverty Bay and the Maori cooking-places near Whangarei. The skull referred here is from the Taupo district. The moa bones found by Mr. Weetman on the Great Barrier Island* Trans. N.Z. Inst., vol. xix., p. 193, and vol. xxii., p. 84. belong to a young chick apparently of this species. They consist of a right metatarsus, a pair of tibiæ, a right femur, and an ischium. The metatarsus (without the distale) is 5.2in. in length and 2.4in. in girth. The tibiæ (without the tibiale) are 9.3in. in length and 2.5in. in girth. The femur is 5.2in. in length and 3.1in. in girth. The ischium is 3in. in length, and has never anchylosed to the other bones of the pelvis. I refer the skull figured in Ext. Birds of N.Z., pl. lii., to this species, because it so closely resembles the skull which we know to belong to D. torosus, and differs much from that of D. potens. The tibia and femur originally described by Professor Owen as those of D. struthioides (Ext. Birds of N.Z., p. 88, †,. 7 and f. Colenso), and afterwards dropped altogether, undoubtedly belong here; while the tibia and femur ascribed to D. gracilis belong really to D. struthioides. Dinornis torosus. D. gracilis, Haast, Trans. N.Z. Inst., vol. i., p. 84, Nos. 7 and 9. Pal. ingens, Hochstetter, New Zealand, p. 187; Jaeger, Reise der “Novara,” Palæ., p. 307, taf. xxv., xxvi. Figures.—Cranium, Ext. Birds of N.Z., pl. lxxxii. (as ingens), Reise der “Novara,” Palæ., pl. xxv., xxvi.; sternum, Ext. Birds of N.Z., pl. xcviii. (as maximus). Restoration, Ext. Birds of N.Z., pl. cxii., fig. 1, and pl. cxiii., fig. 1 (as gracilis), skull incorrect.

Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. No. 1! 12.5 4.0 24.0 5.2 11.5 6.1 No. 2! 25.3 4.8 12.5 6.0 Max. 13.0 4.75 25.3 5.75 12.5 6.6 Min. 12.2 4.0 24.0 4.5 11.5 5.5 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. No. 1! 8.0 8.3 7.0 3.6 3.7 1.7 No. 2! Max. 8.75 8.5 7.2 3.6 3.7 2.0 Min. 7.0 8.0 7.0 2.75 3.7 1.7 Breadth of skull at temporal fossæ, 2.5in. to 2.2in.; at post-frontal processes, 4.1in. to 3.9in. Length of lower jaw, 6.2in. Total length of skull, 7in. Sternum: Length of body, 4.5in.; breadth, 7in.; length of lateral processes 5.5in., of median process 3.5in., but with a notch 0.5in. deep; breadth at anterior border, 7.5in.; across ends of lateral processes, 12in. Length of scapulo-coracoid, 6in. to 5in. Length of ilium, 17in.; breadth of sacrum, 5.5in.; length of ischium, 8in.; depth of pelvis, 6.5in. Distrbution.—Tolerably abundant in the northern part of the South Island as far as Glenmark, but very rare in Otago. No. 1 is a nearly perfect skeleton, in the possession of Mr. R.I. Kingsley, which was found in a cave in the Takaka district, near Nelson. No. 2 is from Glenmark, and is in the Canterbury Museum. This species is intermediate between D. gracilis and D. struthioides, and is more robust than the former. Dr. Hoch-stetter's skeleton, now in the Vienna Museum, undoubtedly belongs here, but the tibiæ belong to D. robustus. This skeleton was collected by diggers and given to the Nelson Museum before Dr. Hochstetter's arrival. The femora were missing. The scapulo-coracoids probably belong to A. didi-formis. Dr. Hochstett says that he himself collected the skull described by Dr. Jaeger, in the same cave, and it probably belonged to the skeleton. The skull belonging to Mr. Kingsley's specimen corresponds so closely in shape and size with it that, no doubt, both belong to one species. Sir R. Owen gives no locality for the skull figured in Ext. Birds of N.Z., pl. lxxxii., but, as he says that he has never received so

good a skull from the North Island, I presume that it came from the South Island. The Hon. W. Mantell informs me that it was not collected by him, and that he knows nothing about its locality. The swelling on the top of the head is greater that in D. gracilis of the North Island. Dinornis struthioides. D. navæ-zealandiæ, Owen, Proc. Zool. Soc. (1843), p. 8. D. struthioides, Owen, Trans. Zool. Soc., vol. iii., p. 235 (1843); and vol. iii., p. 307, cranium. Dinornis struthioides, Reichenbach, Das Nat. Syst. der Vögel. Figures.—Metatarsus, Ext. Birds of N.Z., pl. liv., fig. 4; tibia, pl.lv., fig. 1 (as gracilis) and fig. 2; femur, pl. xxxviii. (as casuarinus), and pl. liv., fig. 1 (as gracilis) and fig. 2; cranium, pl. xvi., figs. 1–4. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 12.0 4.25 No. 1! 11.9 4.25 22.0 4.7 11.25 5.5 No. 2! 23.5 4.5 11.0 4.7 Max. 12.0 5.0 23.5 5.0 11.8 5.8 Min. 11.0 3.0 22.0 4.3 10.0 4.4 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! No. 2! Max. 7.75 6.75 3.25 Min. 2.9 3.0 1.7 Breadth of skull at temporal fossæ, 2.4in. to 2in.; at post-frontal processes, 3.1in. Length of scapulo-coracoid, 4.5in. Length of ilium 14.75in., of ischium 7.25in.; breadth of sacrum, 6in.; depth of pelvis, 6.25in. Distribution.—This is one of the few species of moa which have been found in both Islands. The type specimen is from Poverty Bay, but it was scattered sparingly over both Islands from Whangarei and Mercury Bay in the north to Otago in the south, but was nowhere numerous. It occurs in the swamps of Te Auté, Glenmark, and Hamilton, as well as in the Maori cooking-places in the North Island. No. 1 is an

individual leg in the Wellington Museum, the exact locality of which is not known. No. 2 is from Opito, Mercury Bay, described by Professor Owen as belonging to D. gracilis, Ext. Birds of N.Z., p. 220. The South Island birds were, on the average, rather more slender than those of the North Island, but the difference is very slight. The cranium figured in Ext. Birds of N.Z., pl. xvi., figs. 1.4, is from Poverty Bay. Genus Palapteryx. Palapteryx (part) of Owen; Palapteryx of Reichenbach, not of Haast. Skull moderately depressed, the lambdoidal ridge curved, but the parietals hardly rising above it. The breadth at the squamosals rather more than one and tbree quarter times the height from the basi-temporal. Length from the supra-occipital to the nasals more than the breadth at the squa-mosals. Occipital condyle exposed from above, hidden laterally by the paroccipital processes. Beak about as long as the head, more compressed than in Dinornis; the lower mandible but slightly curved. Plate XV., fig. 3. Sternum with well-marked coracoid pits. The length of the body about three quarters of the breadth. Lateral processes diverging at a small angle from the middle line, short, the breadth across their ends about 1.1 times the breadth of the body. Scapulo-coracoid with a glenoid cavity. Probably, therefore, rudimentary wings were present. Pelvis narrow, and the acetabula set very far back. Greatest breadth at the anti-trochanters about three quarters of the length of the pre-acetabular portion of the ilium. Length of the ilium about 1–8 times the length of the pre-acetabuiar portion. Metatarsus longer than the femur; its length between 2.5 and 3.25 times the girth at the middle of the shaft. Tibia about twice the length of the metatarsus; its length between 4.6 and 5.4 times the girth. Femur more slender than in any other genus; its length from 2.4 to 2.9 times the girth of the shaft. No complete skeleton of this genus has been found, and the different bones are grouped together by inference only. Professor Owen has inferred the existence of a hind-toe, but it has not yet been actually found. The genus is intermediate between Dinornis and Anomalopteryx, but the possession of a wing, the backward positìon of the acetabula, and the slender femora are peculiarities.

Palapteryx dromioides. Dinornis dromioides, Owen, Trans. Zool. Soc., vol. iii., p. 235 (1843), femur. Pal. dromioides, Owen, l.c., vol. iii., p. 307 (1846), metatarsus and tibia. Pal-apteryx dromioides, Reichenbach, Nat. Syst. der Vögel (1850). Figures.—Metatarsus, Ext. Birds of N.Z., pl. xl., fig. 2; tibia, pl. xxxix., fig. 1; femur, pl. xxii., figs. 1, 2, and pl. xxiii., fig. 2; cranium, pl. xlv. (as geranoides), and pl. xlvi., figs. 4–6 (?). Mantell's “Petrifactions and their Teachings,” p. 119. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 10.4 3.75 21.0 4.0 9.6 3.9 Max. 10.4 4.0 21.0 4.7 9.6 4.0 Min. 9.76 3.25 18.4 3.5 9.5 3.8 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. 2.7 2.4 1.3 Min. Breadth of the skull at the temporal fossæ, 1.7in.; at the post-frontal processes, 2.6in. Length of the lower jaw, 5.3in. (?). Length of scapulo-coracoid, 3.25in.; coraco-scapular angle about 125°. Distribution.—The type is from Wanganui sand-dunes, but it has also been found at Poverty Bay and in the Te Aute Swamp. There are two femora in the Wellington Museum from Lyall's Bay. The skull from Wanganui, attributed by Professor Owen to C. geranoides, is certainly too large for that species, while it appears to be too small for D. struthioides; and, as the skulls of S. casuarinus and A. didiformis are known, P. dromioides is the only species left to which it can belong. A small scapulo-coracoid of Dinormis form is in Mr. Hamilton's collection from Te Aute, which, from its size, I refer to this species. It has a distinct glenoid cavity.

Palapteryx plenus. Pal. dromioides, Owen, Trans. Zool. Soc., vol. iv., p. 1 (1850), in part. Figures.—Foot, Ext. Birds of N.Z., pl. li. (as dromioides). Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Max. 10.6 4.0 21.5 4.6 10.0 4.0 Min. 10.25 3.75 20.7 3.8 9.0 3.2 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Max 7.5 6.0 5.25 Min Length of the ilium, 13.5in.; breadth of sacrum, 4.6in.; depth of pelvis, 5.6in. Length of middle toe, 7.3in. Breadth of anterior border of sternum, 6.8in.; across ends of lateral processes, 6.8in. Length of lateral processes, 4in. Distribution.—Found sparingly over the South Island, but everywhere rare. There were a few bones in the swamps at Glenmark and Hamilton. This species is rather larger and stouter than P. dromioides, but with a more slender femur, which approaches that of Megalapteryx hectori. The femur from Glenmark ascribed by Professor Owen to S. rheides no doubt belongs here. The sternum which I have placed with this species also comes from Glenmark. It is on the pattern of the sternum in Syornis, but has coracoid pits; possibly it may belong to D. struthioides, but appears to be too small for that species. The pelvis is from the Manuherikia, Otago, and is in the Wellington Museum. If I am wrong in associating this sternum and pelvis with Palapteryx, the generic characters will, of course, have to be altered. Genus Anomalopteryx. Dinornis (part) of Owen; Anomalopteryx of Reichenbach. Skull large, very convex. Breadth at the squamosals about one and a half times the height from the basi-temporal. Length from supra-occipital to nasals rather more than the

breadth at the squamosals; occipital condyle hidden by the supra-occipital, but exposed laterally: Length of the basi-sphenoid to the end of the rostrum about two and a half times the breadth at the basi-pterygoid processes. The maxillo-jugal bones curved, concave outwards. Temporal fossæ very broad, about equal to the orbits. Beak short, slightly compressed, and rounded at the tip. Length of the premaxillæ to the maxillary suture rather more than one and a quarter times the breadth. Lower mandible strong and nearly straight. Plate XV., fig. 4. Sternum with Well-marked coracoid pits, the length of the body equal to the breadth. Costal processes moderately developed; the lateral processes diverging at an angle of about 20° with the middle line; the posterior median process long Breadth across the ends of the lateral processes, from 1.2 to 1.4 times the breadth of the body. Three pairs of sternal ribs articulating with the sternum. Pelvis narrow, like that of Dinornis; the greatest breadth at the anti-trochanters is rather less than the length of the pre-acetabular portion of the ilium. The position of the ace-tabula varies; the length of the ilium being from 2 to 1.8 times the length of the pre-acetabular portion.. Metatarsus shorter than the femur; its length between 2 and 2.3 times the girth of the shaft. Tibia between 2.1 and 2.2 times the length of the metatarsus; its length between 4.2 and 5.1 times the girth. Length of the femur 1.25 times that of the metatarsus, and between 2 and 2.5 times the girth of the shaft. A hind-toe was present. Anomalopteryx didiformis. Dinornis didiformis, Owen, Trans. Zool. Soc., vol. iii., p. 237 (1843). Anomalopteryx didformis, Reichenbach, Das Nat. Syst. der Vögel. Dinornis parvus, Owen, Trans. Zool. Soc., vol. xi., p. 233 (1882), pl. li.–lvii. Figures.—Metatarsus, Ext. Birds of N.Z., pl. xxvii., figs. 3–6; tibia, pl. xxvi., fig. 3, and pl. xxvii., fig. 3; femur, pl. xx.a, and pl. xxiv.; pelvis, pl. xix., figs. 1–3. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 7.0 3.25 15.5 4.0 8.1 4.25 No. 1! 6.25 2.8 14.0 8.0 3.1 No. 2! 6.1 2.8 12.7 3.0 Max. 7.0 3.4 15.5 4.1 8.8 4.25 Min. 6.0 2.75 12.7 3.0 7.4 3.1

Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 5.5 5.0 3.4 2.6 2.4 1.7 No. 2! Max. 6.0 5.8 4.0 3.0 2.9 1.7 Min. 5.5 5.0 3.4 2.6 2.4 1.6 Breadth of skull at temporal fossæ, 1.7in. to 1.6in.; at post-frontal processes, 3.1in. to 2.9in. Length of scapulo-coracoid, 2in. Length of ilium, 11.5in. to 11in.; breadth of sacrum, 4.25in. to 3.5in.; depth of pelvis, 4.25in. to 4in. Length of middle toe, 4.75in. Total length of skull, 6in. to 5in.; of lower jaw, 5.7in. to 4.7in. Post-frontal process narrow. Distribution.—Found in both Islands, from Whangarei in the north to Southland. In the North Island the bones were tolerably plentiful in the Te Aute Swamp, while the types came from Poverty Bay. It has also been found in the sand-dunes at Lyall's Bay, and in Maori cooking-places at Pataua, near Whangarei. There is a metatarsus in the Canterbury Museum from Horohoro, near Rotorua. In the South Island the bones have been obtained almost entirely from caves, especially in the Nelson District. A few bones were found at Glenmark, and a single tibia at Hamilton. Probably the bird differed in habits from the other species. No. 1 is the type of D. parvus, from Takaka, and No. 2 is a specimen in the Nelson Museum from the same locality. The North Island birds appear to have been, on an average, larger than those of the South Island, and to have had a slightly stouter femur, but there are no differences of specific importance. D. parvus is only a small individual of D. didiformis. I have not seen the scapulo-coracoid; but those obtained by Dr. von Hochstetter from the cave at Collingwood, “scarcely 2in. long” (“New Zealand,” p. 186), probably belong to this species, bones of which occurred in the cave. Certainly they do not belong to the skeleton on which he has placed them. Anomalopteryx antiquus. “Avian Remains,” H. O. Forbes, Trans. N.Z. Inst., vol. xxiii., p. 369. Figures.—Pl. XVII.

Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 5.5 (?) (?) 12.0 3.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Tibia: Breadth at the proximal end, 3.2in.; at the distal end, 2.5in.; at the middle of the shaft, 1.lin.; antero-posterior thickness of shaft, 0.7in. Distribution.—Found at Timaru in a bed of clay of Upper Miocene or older Pliocene age. Mr. Stubbs's collection at Timaru contains two broken tibiæ which are sufficiently perfect to allow measurements of the bone to be taken. One (Pl. XVII., a), about 6in. long, is the distal end of a right tibia, and shows the lateral position of the groove for the tendon of the extensor muscle of the toes, as well as the anterior prominence of the inner condyle, which are so characteristic of the tibia in the Dinornithidæ. The osseous bridge is broken away. The other specimen (Pl. XVII.,b), 8 ½in. long, is the proximal portion of a left tibia which agrees closely in size with the first specimen, so that the total length of the bone can be estimated with considerable accuracy. Both are imbedded in red clay.* I cannot understand Mr. Forbes's remark, with reference to these fossils, that” the largest, nearly 8in. in length, were undoubtedly portions of Dinornis bones of one of the greater forms” (l.c., p. 367). I regret that I am unable to give any measurements of the broken metatarsi in Mr. Miller's collection, as these bones are not now in Timaru, and I have not been able to see them. Consequently the length given for the metatarsus is conjectural. However, photographs of the proximal end of a right metatarsus (Pl. XVII., c) show distinctly the grooved hypotarsal process, with the outer side more prominent than the inner, which is also a character of the Dinornithidæ. The species is about the size of Cela geranoides, but it is very unlikely that it belonged to Cela, because no bones of that genus have ever been found in the South Island. Anoma-lopteryx contains the smallest individuals known in the South Island, and as it is also found in the North Island the genus must have been in existence in the older Pliocene, when the two Islands were joined. This consideration induces me to

refer them to Anomalopteryx. The species makes a near approach to the specimen No. 2 of A. didiformis from Takaka, in the Nelson Museum; but it is smaller, and I think it advisable to keep it separate from that species, partly on this account and partly because of its far greater age. It is, no doubt, the ancestor of A. didiformis. Genus Cela. Dinornis (part) and Palapteryx (part) of Owen; Cela of Reichenbach, not of Mœhring (1752), which is a synonym of Casuarinus. Skull convex, the temporal fossæ very large. Breadth at the squamosals, about 1.6 or 1.7 times the height at the basi-temporal. Length from the supra-occipital to the nasals rather less than the breadth at the squamosals; occipital condyle hidden by the supra-occipital. Ridge between temporal fossæ and supra-occipital narrow. Beak short, slightly compressed and rounded at the tip (but more pointed than in Anomalopteryx); the length of the premaxillæ to the maxillary suture about one and a half times the breadth. Lower mandible nearly straight, and rather slighter than in Anomalo-pteryx. Plate XVI., fig. 5. Sternum with the coracoid pits faintly indicated or altogether absent; length of the body less than the breadth; costal processes well developed; lateral processes diverging at different angles. Pelvis rather broader, in proportion, than in Dinornis, and the acetabula set more forward. The greatest breadth at the anti-trochanters is 1.1 times the length of the pre-acetabular portion of the ilia; the length of the ilium is from 2.4 to 2.5 times the length of the pre-acetabular portion. Metatarsus shorter than the femur; its length between 1.7 and 2.25 times the girth at the middle of the shaft. Tibia about 2.2 times the length of the metatarsus; its length between 3.8 and 4.4 times the girth of the shaft. Femur with a length of 1.2 times that of the metatarsus, and between 1.8 and 2.4 times the girth of the shaft. The presence of a hind-toe has not been proved. This genus contains the smallest species of moa, and is confined to the North Island. Cela geranoides. Palapteryx geranoides, Owen, Trans. Zool. Soc., vol. iii., p. 345 (1848); metatarsus and cranium as of dromioides, not that of geranoides. Figures.—Metatarsus, Ext. Birds of N.Z., pl. lxx., figs. 5, 6; cranium, pl. xxxi., figs. 4, 5, and pl. xlvi., figs. 1–3.

Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 6.0 No. 1! 5.5 3.0 7.0 3.5 Max. 6.1 3.5 13.1 3.25 7.9 4.0 Min. 5.4 2.7 11.7 2.5 6.75 3.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 4.75 Max. 5.5 5.6 2.5 2.2 Min. 5.0 4.75 2.35 1.55 Breadth of skull at temporal fossæ, 1.7in. to 1.65in.; length of ilium, 12.5in. to 12in.; breadth of sacrum, 4.7in. to 4in.; depth of pelvis, 5in. to 4.5in. Distribution..—This was the commonest species in the Te Aute Swamp; and it was also common in the sand-dunes near Wanganui, but rare at Whangarei. The type came from Waingongoro, near Wanganui. No. 1 are specimens in the Otago Museum collected together at Lyall's Bay, Wellington. In Mr. Hamilton's collection from Te Aute there is a broken sternum, intermediate in size between those of C. curtus and A. didiformis, which probably belongs to this species. It is too fragmentary to afford reliable measurements, but it differs from the sterna of both the species just mentioned in having the lateral processes slender, slightly curved, and diverging at a wide angle from the middle line. The cranium, figured on plate xxxi., figs. 4, 5, is from the Bay of Islands; that on plate xlvi., figs. 1–3, is from Wanganui. Cela curtus. Dinornis curtus, Owen, Trans. Zool. Soc., vol. iii., p. 325 (1846). Cela curtus, Reichenbach, Das Nat. Syst. der Vögel. Dinornis oweni, Haast, Trans. Zool. Soc., vol. xii., p. 171 (1886), pl. xxxi., xxxii.* Dromornis australis, Owen, was originally called Dinornis owenii by Dr. Krefft, but the name was, I believe, only published in a Sydney newspaper. Figures.—Metatarsus, Ext. Birds of N.Z., pl. xl., fig. 6, and pl. lxxxvii., figs. 7–10; tibia, pl. xxxix., figs. 1–5; femur,

pl. lxviii, figs. 5, 6 (as geranoides); pelvis, pl. xx.a (as didiformis). Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 5.0 2.8 11.25 2.75 No. 1! 4.4 2.3 9.6 2.4 5.65 2.6 Max. 5.25 3.0 11.6 3.1 6.5 3.5 Min. 4.4 2.2 9.5 2.25 5.5 2.6 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 4.0 3.5 2.4 1.4 Max. 2.7 Min. Breadth of skull at temporal fossæ, 1.65in.; at post-frontal processes, 2.65in. The skull is as large as that of C. geranoides, but differs from that species in the backward slope of the supra-occipital, the great prominence of the lambdoidal ridge, as well as the greater flattening of the parietals. The sternum has the lateral processes diverging at a small angle, as in Syornis. Distribution.—The type is from Poverty Bay. It was common in the Te Aute Swamp, and the most abundant of all the species at Whangarei; taking the place in the north of C. geranoides in the south. No. 1 are the measurements of the type specimen of D. oweni, from Whangarei, now in the Auckland Museum. I regret that I cannot maintain D. oweni as a separate species. It is only a small individual of C. curtus, with the same proportions and the same geographical distribution. Sir J. von Haast says that the femur is stouter and the metatarsus more slender in curtus than in oweni; but my measurements do not bear this out, and the supposed anatomical differences between them are only individual variations, which may be found in almost every species of moa. If oweni is to be separated from, curtus, then, for the sake of uniformity, most of the species should be split into two, for they show quite as wide a range of variation.

Genus Mesopteryx.* Greek μεσóoς, middle. Dinornis (part) of Owen. Skull convex, angled behind. The vertical ridge of the supra-occipital broadened out into a triangular tuberosity. Breadth at the squamosals one and a half times the height from the basi-temporal. Length from the supra-occipital to the nasals rather greater than the breadth at the squamosals. Temporal fossæ narrow and with a median elevation. Occipital condyle hidden by the supra-occipital, slightly exposed laterally. Orbits small. Beak slender, shorter than the head, much compressed and pointed at the tip. Length of the premaxillæ to the maxillary suture, not quite one and a half times the breadth. Lower mandible slender. Plate XVI., fig. 7. Sternum without coracoid pits. Length of the body about three quarters of the breadth. Costal processes very large; lateral processes narrow, slightly curved outwards, diverging at an angle of about 20° from the middle line; median process long and not notched. No scapulo-coracoid. Pelvis moderate; the breadth at the anti-trochanters about 1.2 times the length of the pre-acetabular portion of the ilia; length of the ilium about 2.3 times that of its pre-acetabular portion. Metatarsus shorter than the femur; its length between 1.7 and 2 times the girth at the middle of the shaft. Tibia about 2.25 times the length of the metatarsus, its length between 3.7 and 4 times the girth of the shaft. Femur about 1.35 times the length of the metatarsus; its length between 1.8 and 2.2 times the girth of the shaft. A hind-toe was present, and the legs were feathered down to the toes. Mesopteryx didinus. Dinornis didiformis, Haast, Trans. N.Z. Insts, vol. i., p. 83, Nos. 5 and 6. Dinornis huttonii, Owen, Ext. Birds of N.Z., p. 430 (1879). Dinornis didinus, Owen, Trans. Zool. Soc., vol. xi., p. 257 (1882). Figures.—Cranium, Ext. Birds of N.Z., pl. lxxviii. (as casuarinus). Restoration, pl. cxi. (as didiformis). Metatarsus. Tibia. Femur Length. Girth. Length. Girth. Length. Girth. Type 6.75 3.9 15.25 Max. 7.2 4.0 15.25 4.1 9.2 4.9 Min. 6.5 3.25 14.5 3.6 8.9 4.0

Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. 6.8 8.0 4.25 2.7 2.5 1.7 Min. 6.0 7.25 2.7 2.3 1.5 Breadth of the skull at the temporal fossæ, 1.9in. to 1.7in.; at the post-frontal processes, 2.8in. to 2.7in. Length of the lower jaw, 4.25in. to 4in. Total length of the head, about 5in. Length of the ilium, 15.5in. to 14in.; of the ischium, 7.5in. to 6.5in.; breadth of the sacrum, 7in. to 6.2in.; depth of the pelvis, 6in. to 5.5in. Distribution.—Found in both Islands, but rare in the North. The type is from a cave near Queenstown; and bones have also been found in the peat-beds at Hamilton and Glenmark, and in the caves at Takaka. In the North Island, a few bones have been obtained in the Te Aute Swamp and at Poverty Bay. This species has been confounded with A. didiformis, but is stouter in all bones of the leg. It is also sharply marked off from S. casuarinus by the length of the tibia, although no such break exists with the metatarsi. The skull is intermediate between Cela and Syornis, but stands alone in the slenderness of the beak. Sir R. Owen gave the name D. huttonii to this species on the strength of my statement that it was different from A. didiformis, but without any description. Afterwards he described it under the name of D. didinus; and I think that the name that accompanied the description ought to take precedence. Genus Syornis. Dinornis (part) of Owen; Syornis and Emeus of Reichenbach; Meionornis, Palapteryœ (part), and Euryapteryx (part) of Haast. Skull convex, rounded behind, the supra-occipital sloping forward to the lambdoidal ridge. Breadth at the squamosals, from 1.6 to 1.7 times the height from the basi-temporal. Length from the supra-occipital to the nasals equal to, or more than, the breadth of the squamosals. Length of the basi-sphenoid from the basi-occipital to the end of the rostrum three to three and a half times the breadth of the basi-pterygoid processes. Occipital condyle just hidden or slightly exposed both vertically and laterally. Temporal fossæ broad. Beak shorter than the head, moderately curved, much compressed and pointed at the tip; length of the premaxillæ to

the maxillary suture about one and three quarter times the breadth. Lower mandible strong. Plate XVI., fig. 6. Sternum without coracoid pits. Length of the body, from 1.1 to 0.75 times the breadth. Costal processes large; the lateral processes nearly straight, diverging at an angle of about 20° from the middle line; the median posterior process often long, rarely notched at the end. Breadth across the ends of the lateral processes, about 1.1 times the breadth of the body below the costal region. No scapulo-coracoid. Pelvis moderately broad; the breadth at the anti-trochanters about 1.25 times the length of the pre-acetabular portion of the ilia. Length of the ilium, between 2.2 and 2.5 times that of its pre-acetabular portion. Metatarsus shorter than the femur; its length between 1.6 and 2 times the girth at the middle of the shaft. Tibia from 2.1 to 2.25 times the length of the metatarsus; its length between 3.7 and 4.4 times the girth of the shaft. Femur about 1.25 times the length of the metatarsus; its length between 1.9 and 2.1 times the girth of the shaft. The presence of a hind-toe has been proved in S. casuarinus. The tracheal rings are ossified, thick, circular, deep, and rough. Ext. Birds of N.Z., pl. xciii., figs. 13–32. Syornis rheides. Dinornis rheides, Owen, Trans. Zool. Soc., vol. iv., p. 1, (1850), metatarsus only. Not D. rheides, Ext. Birds of N.Z., pl. cix. Metatarsus. Tibia. Femur Length. Girth. Length. Girth. Length. Girth. Type 9.0 4.5 Max. 9.75 6.0 22.5 5.75 12.3 6.5 Min. 8.8 4.5 20.0 4.9 11.5 6.0 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Max. 9.5 12.0 Min… 8.5 10.5 Length of ilium, 22.5in. to 19.5in.; of ischium, 9.7in.; breadth of sacrum, 9.7in. to 9in. Length of lower jaw, 6.75in. to 6.2in.

Distribution.—From Motunau and Glenmark, in Canter-bury, to Hamilton and Shag Point, in Otago. It was commonest at Hamilton, and rare in the sand-dunes at Shag Point. This species has been much misunderstood owing to Professor Owen having associated a tibia ofS. crassus and a femur of P. plenus with the metatarsus.*The foot figured by Professor Owen in Ext. Birds of N.Z., pl. 1., also appears to be that of P. plenus. Sir J. von Haast confused it with E. gravis, probably on account of the supposed length of the tibia, and the photograph on plate cix. of the “Extinct Birds of New Zealand” represents this mixture. The femur is difficult to distinguish from that of D. torosus on one hand and E. elephantopus on the other. A complete leg has not yet been found. Syornis crassus. Dinornis crassus, Owen, Trans. Zool. Soc., vol. iii., p. 307 (1846), and vol. iv., p. 159, cranium. Emeus crassus, Reichenbach, Das Nat. Syst. der Vogel. Palapteryx crassus, Haast. Figures.—Metatarsus, Ext. Brids of N.Z., pl. xl., figs. 4, 5; tibia, pl. xxxix., fig.2 (as casuarinus); cranum, pl. lix. and lxxvi. (as elephantopus). Restoration, pl. cxii., fig. 2 (as casuarinus). Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 8.5 4.7 No. 1! 8.6 5.2 19.5 4.9 11.6 6.0 No. 2! 8.6 4.4 18.6 4.6 10.5 5.6 Max. 8.75 5.5 19.9 5.2 11.6 6.0 Min. 8.2 4.0 18.5 4.4 10.5 5.6 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! No. 2! Max. 8.5 10.5 6.75 4.0 3.45 2.0 Min. 7.75 9.75 6.0 3.7 3.35 2.0 Breadth of skull at temporal fossae, 2.3in. to 2.2in.; at post-frontal processes, 3.6in. to 3.5in. Length of the lower jaw, 5.8in. to 5.7in. Total length of the head, 6.5in. Postfrontal

process broad; breadth of the temporal fossae, about two-thirds that of the orbits. The post-axial tarsal bone is about 2in. in length and 1in. in greatest breadth. The length of the ilium is 20.5in: to 18.5in.; of the ischium, 9.5in. to 8.25in.; breadth of the sacrum, 8.75in. to 8.5in.; depth of the pelvis, 8in. to 7in. Distribution.— Very common in the peat-beds at Hamilton and Glenmark; also found in Maori cooking-places at Awa-moa and Shag Point. The femur with fragments of muscle which accompanied the neck with integument, found in Earns-cleugh Cave, belongs to this species.* Trans. N.Z. Inst., vol. iv., p. 112. The type is from Waikouaiti. The specimen No. 1 is from Glenmark; No. 2 is from a leg in the Wellington Museum the locality of which is unknown. This species has, like the last, been misunderstood. Professor Owen gave it its name before he had seen bones of E. elephantopus; and the tibia and femur which he associated with the metatarsus belonged to a smaller species. These gave the idea that the bird was very robust; and Sir J. von Haast, as well as myself, placed it close to E. elephantopus. The skull ascribed to this species by Sir R. Owen has been proved, by the skeleton in the Otago Museum, to belong to E. elephantopus, and consequently the skull placed on that species must come here. The femora are difficult to distinguish from those of D. struthioides on the one hand and E. ponderosus on the other. Syornis casuarinus. Dinornis casuarinus, Owen, Trans. Zool. Soc., vol. iii., p. 307 (1846), metatarsus only; and vol. vii., p. 115, sternum. Syornis casuarinus, Reichenbach, Das Nat. Syst. der Vögel (1850). Meionornis casuarinus, Haast, Trans. N.Z. Inst., vol. vii., pp. 54, 91 (1874). Figures.—Metatarsus, Ext. Birds of N.Z., pl. xl. fig. 3; femur, pl. xxiii., fig. 1; cranium, pl. lxxv. (as rheides); sternum, pl. xlviii., and pl. lxxiii., lxxiv. (as rheides). Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 8.0 4.2 No. 1! 7.9 4.6 18.0 4.2 10.3 5.4 No. 2! 7.25 4.5 18.0 4.1 10.5 5.2 Max. 8.0 5.0 18.25 4.75 10.3 5.75 Min. 7.25 3.6 16.0 3.3 9.3 4.25

Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 6.0 7.75 5.0 No. 2! Max. 7.0 9.5 5.5 4.0 3.3 2.1 Min. 8.8 5.0 3.3 3.2 1.9 Breadth of skull at temporal fossæ, 2in. to 1.9in.; at post-frontal processes, 3.75in. to 3.55in.; length of lower jaw, 4.5in.; total length of head, 5in. Temporal fossæ narrow, the breadth about one-third that of the orbits. Length of the ilium, 17.5in. to 16in.; of the ischium, 8.7in. to 7in.; breadth of the sacrum, 8in. to 7in.; depth of the pelvis, 6.2in. to 5.6in. Length of the middle toe, 6in. Distribution.—Found in both Islands, but rare in the North and common' in the South. The type is from Wai-kouaiti, in Otago. No. 1 is a specimen in the Otago Museum found at Dunedin;* Hector, Proc. Zool. Soc., 1865, p. 749. No. 2 is a specimen in the Wellington Museum, the locality unknown. It was very abundant in the peat-beds of Glenmark and Hamilton, and occurs also in the Maori cooking-places at Shag Point. A few bones were found in the Te Aute Swamp, and it also occurs in the Maori cooking-places at Wanganui and Whangarei. The North Island birds agree with those from. Glenmark Swamp in having a rather shorter tibia than the birds from Otago; while the rather shorter tibia than the birds from Otago; while the rather older birds from Motunau agree in size with the southern variety. This difference is sufficient to distinguish two varieties, but is not of specific importance. In Otago the length of the tibia is between 18.25in. and 17in., while in Glenmark and the North Island it is between 17.8in. and 16in. A sternum in the Wellington Museum from Waipawa, Hawke's Bay, enables us to identify the sterna found in the South Island. The cranium figured in Ext. Birds of N.Z., pl. liii., fig. 6, may belong here, but is rather large. The femora are difficult to distinguish from those of E. gravis. Genus Euryapteryx. Dinornis (part) of Owen; Palapteryx(part)and Eury-apteryx (part) of Haast.

Skull rather small, rounded behind,' and moderately convex. Breadth at the squamosals from 1.5 to 1.6 times the height at the basi-temporal. Length to the nasals either greater or less than the breadth of the squamosals. Length of the basi-sphenoid from the basi-occipital to the end of the rostrum rather less than three times the breadth of the basi-pterygoid processes. Temporal fossæ narrow and deep. Posterior margin of tympanic cavity straight. Occipital condyle hidden or slightly exposed both vertically and laterally. Beak very short and stout, slightly compressed and rounded at the tip. Length of the premaxillæ to the maxillary suture less than the breadth. Plate XVI., fig. 8. Sternum without coracoid pits, or sometimes faintly marked. Length of the body about one-half of its breadth. Costal processes small; lateral processes short, nearly straight, diverging at an angle of about 35° from the middle line; median process short, rarely notched. Breadth across the ends of the lateral processes, from 1.8 to 2.1 times the breadth of the body below the costal region. Three pairs of sternal ribs articulating with the sternum. No scapulo-coracoid. Pelvis very broad, and the acetabula set far forwards. The breadth at the anti-trochanters is one and a half times the length of the pre-acetabular portion of the ilia. Length of the ilium, from 2.7 to 2.9 times the length of the pre-acetabular portion. Metatarsus much shorter than the femur; its length between 1.3 and 1.5 times the girth of the shaft in the middle. Tibia from 2.2 to 2.4 times the length of the metatarsus; its length between 3 and 3.7 times the girth of the shaft. Femur about 1.3 times the length of the metatarsus; its length between 1.8 and 1.5 times the girth of the shaft. The presence of a hind-toe has been proved in E. elephan-topus and E. ponderosus. The tracheal rings are ossified, elliptical, and smooth. Ext. Birds of N.Z., pl. xciii., figs. 1–12. This genus contains the most robust of all the moas, and is confined to the South Island. Euryapteryx elephantopus. Dinornis elephantopus, Owen, Trans. Zool. Soc., vol. iv., p. 149 (1856), and p. 159; Buller, Birds of N.Z., 2nd., vol. i., p. xxxii., and vol. ii., p. 333; Haast, Trans. N.Z. Inst., vol. i., p. 80, No. 12. Figures.—Metatarsus, Ext. Birds of N.Z., pl. lvii.; tibia, l.c..., pl. lvi., fig. 4; femur, pl. lvi., fig. 1; cranium, l.c.,

pl. Ixxvii., and pl. cxiv., fig. 1 (as crassus); sternum, pl. lxxii. Restoration, pl. cviii., and pl. cxiii., fig. 2 (as crassus). Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 9.25 6.5 24.0 6.4 13.0 7.75 No. 1! 9.1 6.0 21.1 5.6 11.8 6.8 No. 2! 9.8 6.8 22.8 6.3 12.8 7.7 No. 3! 9.5 6.8 23.0 6.5 13.0 7.6 Max. 9.9 7.2 24.0 6.5 13.0 8.25 Min. 9.0 5.9 20.8 5.7 11.75 6.4 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! 7.5 11.5 8.0 No. 2! No. 3! Max. 7.5 12.0 8.8 3.0 3.1 1.9 Min. 6.5 11.5 8.0 2.85 2.6 1.8 Breadth of skull at temporal fossæ, 2.2in. to 2.1in.; at post-frontal processes, 3.45in. to 3in. Length of lower jaw, 5.25in. to 4.8in. Total length of the head, about 5.5in. Length of ilium 19in.; of ischium 9in.; breadth of the sacrum, 10.25in. to 10in.; depth of pelvis, 7in. to 6in. Post-axial tarsal with a length of from 2.5in. to 2in., and greatest breadth 1.25in. to 1in. Length of middle toe, 8.5in. to 7in. Distribution. — South Island, from Takaka, near Nelson, to Otago; the type is from Awamoa, near Oamaru. Most abundant in Otago, and comparatively rare at Glen-mark. No. 1 is a specimen from a cave in the Waitaki Valley, now in the Otago Museum; Nos. 2 and 3 are from Glenmark. The cranium figured in Ext. Birds of N.Z., pl. cxiv., fig. 1, was obtained by me in the sandhills at the mouth of the Shag River, and sent to the British Museum. The femur from Waikouaiti ascribed by Professor Owen to S. crassus belongs here, as also do the skulls described by him under the same name.

Euryapteryx ponderosus. Dinornis gravis (?), Hutton, Trans. N.Z. Inst., vol. vii., p. 275. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. No. 1! 8.25 6.0 18.5 5.25 No. 2! 8.0 5.0 18.6 5.4 10.0 5.6 Max. 8.9 6.75 20.6 5.9 11.0 6.8 Min. 8.0 5.0 18.5 5.2 10.0 5.4 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. No. 1! No. 2! Max. 6.0 10.0 7.5 2.8 2.8 1.85 Min. 7.0 2.5 2.7 1.7 Breadth of the skull at temporal fossæ, 2in. to 1.9in.; at the post-frontal processes, 3.1in. to 2.95in. Length of lower jaw, 4.8in. to 4.5in. Total length of the head, 5.5in. Length of the ilium, 16.5in.; breadth of the sacrum, 8.25in.; depth of the pelvis, 6in. The post-axial tarsal has a length of 2in. to 1.75in., and a greatest breadth of 0.9in. to 0.8in. Length of the middle toe, 6in. Distribution.—Throughout Otago and Canterbury. Not known from the Nelson District. It was the most abundant species in the Maori cooking-places at Shag Point, where also the skulls were abundant; and it occurs in the Pleistocene deposits at Motunau. No. 1 is from Hamilton; No. 2 is the measurements of a pair of legs in the Wellington Museum, the locality of which is not known. The skull of this species can be distinguished from that of the last by the processes at the hinder angles of the basisphenoid, which are higher and rounder in ponderosus, flatter and more elongated in elephantopus. From E. gravis the skull is distinguished by the parietals which are convex in all directions in ponderosus, but concave longitudinally in gravis, the post-frontal elevation on the skull being more conspicuous in that species. But there are intermediate forms. The tibia and femur appear to be well marked off from those of E. gravis, but all the bones pass into those of E. elephantopus. I have separated it from this last species partly on account

of its geographical distribution, and partly because a range of variation of two inches in the metatarsus is much more than what has been allowed for the species of Syornis and Cela. It will be seen also that there are three types of skull in Euryapteryx. Euryapteryx gravis. Dinornis gravis Owen, Trans. Zool. Soc., vol. viii., p. 361, 1872 (metatarsus and tibia only). D. sp., Haast, Trans. N.Z. Inst., vol. i., No. 17. Figures—Metatarsus, Ext. Birds of N.Z., pl. xlii.a; tibia, l.c., pl. xlii., figs. 1–3; cranium, l.c., pl. lxxxi. (not fig. 5). Restoration, pl. cx. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 7.75 5.0 16.6 5.0 No. 1! 7.75 5.0 15.75 4.75 Max. 7.8 5.5 17.75 5.1 9.5 5.3 Min. 7.0 4.5 15.7 3.3 9.0 4.7 Pelvis. Sternum. Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type No. 1! Max. 4.5 7.0 6.5 2.5 2.7 1.7 Min. 2.25 2.5 1.6 Breadth of the skull at temporal fossæ, 1.85in. to 1.75in.; at post-frontal processes, 4.1in. to 3.1in. Length of lower jaw, 4.4in. to 4.3in. Total length of the head, 5in. The temporal fossæ are broader than in the other species, being about one-half of that of the orbits. Length of the ilium, 12in.; breadth of the sacrum, 6in.; depth of the pelvis, 5in. Distribution.—The type is from Kakanui, near Oamaru. It was common in the peat-beds at Hamilton and Glenmark, and in the Pleistocene deposits at Motunau. It also occurs at the Maori cooking-places at Shag Point. No. 1 is from Hamilton.

Euryapteryx pygmæus. Metatarsus. Tibia. Femur. Length. Girth. Length. Girth. Length. Girth. Type 6.0 4.0 13.5 5.0 7.5 4.5 Pelvis. Sternum Skull. Length. Breadth. Breadth. Length. Breadth. Height. Type Metatarsus: Breadth at proximal end, 2.65in.; at middle of shaft, 1.55in.; at distal end, 3.4in.; antero-posterior thickness of shaft, 0.9in. Distribution.—This species is founded on a pair of metatarsi in the Nelson Museum from Takaka. With them I have associated a tibia, and a femur from Otago, which were exhibited at the Dunedin Exhibition in 1865, the measurements of which are given by Sir James Hector in Proc. Zool. Soc., 1865. It appears to have been a very rare species. History of the Moas. The history of the moas presents several interesting problems not yet completely solved, prominent among which are their origin, their advent in New Zealand, their development, and the date of their extinction. Origin of the Moas. The moas belong to that group of birds called “Ratitæ” by Merrem (1813), to which also belong the ostrich, the rhea, the emu, the cassowary, and the kiwi. All of these have small, or even rudimentary, wings, and loosely-arranged feathers; while all but the kiwi are birds of large size. Whether the Ratitæ—or struthious birds—form an offshoot from the Carinatæ—or flying birds—or whether the Carinatæ are descended from the Ratitæ, is a question which has been much discussed. But the writings of Professor T. J. Parker on Notornis and on Apteryx* Trans. N.Z. Inst., vol. xiv., p. 254; N.Z. Journal of Science, 1891, pp. 3 and 66; Phil. Trans., vol. clxxxii., p. 25. have, I think, conclusively settled it in favour of the first view. I think, conclusively settled it in favour of the first view. Indeed, the structure of the rudimentary wings, and the cellular texture of some of the

bones, are by themselves sufficient evidence to show that the remote ancestors of the Ratitæ could fly. Palæontology confirms this opinion, because the oldest known bird—Archæo-pteryx, of the Jurassic period—had well-developed wings. With the exception of some peculiarities in the base of the skull, all the other characters which unite the Ratitæ into one group, including their great size, are merely adaptations to a terrestrial life, and are of little value as evidence of affinity. Even the characters of the skull are not confined to the Ratitæ, but are also found in the tinamous and Opisthocomus, which are carinate birds. The living Ratitæ form four well-defined families— 1. Struthionidæ, containing the two-toed ostrich of Africa. 2. Rheidá, containing the rhea of South America. 3. Casuridæ, including the emu and cassowary of Australia and Melanesia. 4. Apterygidæ, containing the kiwi of New Zealand. The moas— Dinornithideæ—come between the Casuaridæ and the Apterygidæ, connecting them together. Contemporaneous with the moas were Dromormis and extinct species of emu in Australia, a large extinct species of rhea in Brazil, and Æpyornis in Madagascar. The last of these was a struthious bird, but it is doubtful where it should be placed. Certainly it was not closely allied to the moas, although it resembled some of them in its gigantic size, and in having a hind-toe. In the early part of the Pliocene period an ostrich lived in northern India, as also did another with three toes, perhaps allied to the rhea; and here we come to the earliest certain traces of struthious birds outside of New Zealand. In the Lower Eocene, Dasornis, Megalornis, and Macroornis, of Europe, * Gastornis is said both by A. Milne-Edwards and E. T, Newton to be related to the Anatidæ (Trans. Zool. Soc., vol. xii., p. 143); but Mr. Lydekker places it doubtfully in the Ratitæ (Knowledge, 1891, and Nicholson's “Palæontology,” vol. ii.). as well as Dia-tryma, of Mexico, are thought by some to belong to the Ratitæ. The fragment of the skull of Dasornis is said by Professor Owen to resemble the skulls of Rhea and Struthio more than that of Dinornis. The fragmentary shaft of the tibia of Megalornis is compared to that of the emu by Professor Seeley.† Quar. Jour. Geol. Soc., vol. xxx., p. 708. Professor Cope thinks that Diatryma was allied to the rhea, while Mr. Lydekker thinks it may be identical with Gastornis. Great size may go either with aquatic or with terrestrial habits, and there is as yet no conclusive proof that any of these birds belong to the Ratitæ; and certainly there

is no evidence for supposing any of them to be the ancestors of the moas. Mesembriornis, of South America, appears to be related to the Anatidæ. In the Cretaceous period all the birds were very different from living ones, and, although flying birds and flightless birds existed even then, the flightless birds were adapted for swimming, while the Ratitæ are a group specially adapted for terrestrial life. It is true that the Cretaceous birds had skulls, so far as we know, on the present Ratite type; but that pattern appears to have been at the time common to all birds. There were no distinct Carinatæ and Ratitæ, only the Protocarinatæ of Professor T. J. Parker. The Ratitæ must have branched off from the Proto-carinatæ at an early date, but probably not before the close of the Cretaceous period. The Grallæ of Bonaparte, together with the bustards, form another group, also adapted for terrestrial life, which branched off from the Carinatæ at a later period. Advent of the Moa in New Zealand. That the moas have been a long time in New Zealand is evident. Numbers of bones have been obtained from old Maori cooking-places,* Murison, Trans. N.Z. Inst., vol. iv., p. 122; Haast, l.c., vol. vii., pp. 86 and 91; Thorne, l.c.., vol. viii., p. 83; Hutton, l.c., vol. viii., p. 103; Robson, l.c., vol. viii., p. 95, and vol. ix., p. 279; Mantell, l.c., vol. xxi., p. 440; G. A. Mantell, “Petrifactions and their Teachings,” p. 101. which are, of course, recent; and still greater numbers have been found in swamps and caves, some of which are of Pleistocene age. At the Patangata Swamp, Te Aute, near Napier, the bones were partly in a stiff blue clay and partly in an old forest-bed lying on the clay, and were covered by 8ft. or 10ft. of silt;† Hamilton, Trans. N.Z. Inst., vol. xxi., p. 314; Park, Rep. Geol. Expl., 1887–88, p. 88. while others occurred lying on the surface of the drained lagoon, and may have been younger. The swamp at Glenmark, which had to be drained, may possibly belong to the Recent period, but below the swamp there is a series of beds of river-shingle, peat, and silt, covered by the loess deposit so common in Canterbury, which is no doubt Pleistocene, and from this peat-bed a large number of moa-bones have been obtained.‡ Haast, “Geology of Canterbury,” p. 442. At Hamilton, in central Otago, the principal deposit of bones was a small dry peat basin excavated out of a bed of clay, and below the clay, which was 6ft. thick, there was another small peat basin, also full of moa-bones.§ Booth, Trans. N.Z. Inst., vol. vii., p. 123, and vol. ix., p. 365; Hutton, Quar. Jour. Geol. Soc., vol. xli., p. 213. Considerable changes have taken place in the physical geography of the country since these peat basins were

formed, and undoubtedly they must be referred to the Pleistocene. Bones have been found at Motunau, in clay and lignite, which Sir J. Hector and Mr. A. McKay consider to be early Pleistocene, and of the same age as the oldest bones at Glen-mark. These beds were covered by 10ft. to 60ft. of alluvial deposits. Mr. McKay has also collected broken moa-bones and fragments of moa egg-shell from the Pleistocene beds at Gore Bay. These beds form a series of angular gravels and silts, with marine shells in the lower portion, passing upwards into rounded gravels, clays, and loam. The moa-bones were obtained from about the middle of the formation.∗∗ Trans. N.Z. Inst., vol. xiv., pp. 410 and 540; Rep. Geol. Expl., 1882, p. 74. Moa-bones also occur in the loess which wraps round Banks Peninsula and covers the low hills at Timaru and Oamaru,†† Hutton, “Geology of Otago” (1875), pp. 71, 72. as well as in the higher alluvia of the rivers. A few bones are reported by Dr. Haast from the ancient moraines in Canterbury and Otago,‡‡ Quar. Jour. Geol. Soc., vol. xxi., p. 134; Trans. N.Z. Inst., vol. iv., p. 68; Geol. Canterbury, pp. 380 and 437. which are probably of Pliocene age. Near Napier moa-bones have been found by Mr. Aug. Hamilton associated with newer Pliocene marine shells;§§ Trans. N.Z. Inst., vol. xxi., p. 312. and the footprints discovered by Mr. D. Millar in the sandstone at Gisborne¶¶ Williams, Trans, N.Z. Inst., vol. iv., p. 122; Gillies, l.c., p. 127; Owen, Ext. Birds of N.Z., p. 451, pl. cxvi. are also probably newer Pliocene. Mr. S. H. Drew has found bird-bones, probably belonging to the moa, in the Pliocene brown sands near Kai-iwi, between Wanganui and the mouth of the Waitotara.∥∥ Park, Rep. Geol. Expl., 1886–87, p. 63. Undoubted moa remains have been found by Mr. F. W. Stubs and Mr. Miller in a bed of clay at Timaru, which overlies gravel and underlies a lava-stream; and with them was associated a femur supposed by Mr. H. O. Forbes to belong to the living Apteryx australis.∗∗∗∗ Trans. N.Z. Inst., vol. xxiii., p. 367. These lavas have hitherto been considered as Miocene, or even Eocene, in age. They can be traced for about twelve miles in a westerly direction, gradually rising until they form the summit of Mount Horrible (1,138ft.); and here they end abruptly in a precipice overlooking the valley of the Pareora River. Below the lava-streams, and separated from them by a thickness of 300ft. or 4000000ft. of sandy and tufaceous beds, are older lava-streams, which on the south-west side of Mount Horrible lie on limestone of Oligocene age. In 1865 Sir J. von Haast though that certain beds, with marine fossils belonging to the Pareora series

(Miocene), found in the immediate neighbourhood, were the equivalents of the tufaceous beds, although he did not succeed in finding the fossiliferous beds actually cropping out between the lava-streams.* “Report on the Geological Formation of the Timaru District,” pp. 10 and 12. But in 1872 he says that the limestone at Mount Horrible (Weka Pass limestone) is “subdivided by several coulées of the anamesites;”† Rep. Geol. Expl., 1871–72, p. 25. and in his “Geology of Canterbury” (1879) he refers the Timaru volcanic rocks to the Oamaru (Eocene) formation. Writing of the section at Mount Horrible, he omits his former statement that lavas are interbedded with the limestone, and says, “Upon the calcareous greensands forming the upper bed of the Oamaru formation, and which is often so rich in carbonate of lime that it can be used for the lime-kiln, a bed of volcanic tufa reposes, sometimes changed into an agglomerate, after which the first lava-stream appears. Then follow a number of tufaceous beds with some smaller lava-streams between them, till the uppermost coulée is reached, having, like the lowest one also, a thickness of about 50ft. The anamesitic rock is extensively quarried near Timaru, and forms a valuable building-stone for that town” (l.c., p. 314). In 1876 Mr. A. McKay, of the Geological Survey, found in this locality blue sandy beds with Miocene marine fossils, overlain conformably by grey sands, lignite, and gravels; and he says that the volcanic rocks of Timaru are associated with the Pareora (Miocene) bed.‡ Rep. Geol. Expl., 1876–77, pp. 54 and 66. A few years later, relying on Dr. von Haast's statement that lavas were actually inter-stratified with the calcareous rocks, he considered that these must be older than the lavas of Mount Horrible and Timaru, which, he says again, belong to the Pareora formation, § Rep. Geol. Expl., 1879–80, p. 78. and form part of the latest display of volcanic energy in this Island. But, as Dr. Haast does not mention this supposed intercalation of calcareous rocks and lava - streams in his “Geology of Canterbury,” we must suppose it to have been a mistake, and consequently there is no reason to regard any of the volcanic rocks near Timaru as of Eocene age. That these lavas were not younger than Miocene was therefore the opinion of the only two geologists who had examined the locality. But last year Mr. H. O. Forbes published his opinion that they were probably of “newer Pliocene or even Pleistocene age.”∥ Trans. N.Z.Inst., vol. xxiii., p. 372. This conclusion was not arrived at by a re-examination of the sections at Mount Horrible and the Pareora. Mr. Forbes merely went to a quarry near

Timaru and, “with little doubt,” identified a “rough red shingle,” which he did not even see in sitû, with the gravels of the “alluvial fans” of the Canterbury Plains. He also identified the bone-bed itself, which lies above the “rough red shingle” and below the lava-stream, with the silt or loess which lies on the top of the lava-stream. Mr. Forbes's opinion on the subject cannot therefore carry any weight; it is merely another instance of the mistake, so often made, of correlating rocks by lithological resemblances, and overlooking the far more important evidence to be derived from the changes which took place between the deposition of the two beds. The moa-remains belong to a species (A. antiquus) which has not been found elsewhere, but if it had been correct that bones of a living species of kiwi occurred with them it would have been strong evidence in favour of the bed being younger than Miocene. A broken femur, only partially cleared from the matrix in which it is imbedded, cannot furnish very conclusive evidence; and Mr. Forbes gives no description of the bone, neither does he say on what characters he relies as proving it to be the femur of Apteryx. Certainly his drawings do not bear out his statement, for they show a bone differing materially from the femur of Apteryx australis. The shaft is too straight and too stout; the head too convex; the neck too transverse to the axis of the bone, and far too much constricted. At the distal end the inner condyle projects inward too much, and the outer one too much downward, thus making the articular surface too oblique to the shaft. In several characters the drawings resemble the femur of Aptornis more than that of Apteryx, and I cannot accept Mr. Forbes's drawings, unaccompanied as they are by any description, as a proof that the bone belongs to Apteryx at all, much less to A. australis. The gravels under the bone-bed belong, no doubt, to the series of sands and gravels which Mr. McKay has shown to lie conformably on beds with marine Miocene fossils. Whether these gravels are Upper Miocene or older Pliocene it is at present impossible to say; but we may safely assume that the lavas are much older than Pleistocene, because since they were erupted the amount of denudation has been enormous, and all traces of the place from whence they came have been swept away. Lastly, the Hon. W. Mantell, in 1849, found in a septarium near Hampden, Otago, a fragment of a large bone 1 ½in. in diameter, which was said by high authorities in England to belong to a bird.* Quar. Jour. Geol. Soc., vol. vi., p. 326. These septaria are, I think, of Lower Miocene age, as they occur in sandy clays containing in the close neighbourhood Aturia zic-zac, Ancillaria, Cominella, as

well as other fossils, including at least three living species of Mollusca,* Trans. N.Z. Inst., vol. xix., p. 426. and, if this fragment is part of a moa-bone, it is the oldest at present known. Doubts have been expressed as to the ornithic nature of this fragment, and it has been suggested that it may be reptilian. Mr. H. O. Forbes has even gone so far as to say that, “as reptilian bones have since been obtained from the same horizon, it was probably of this nature.”† Trans. N.Z. Inst., vol. xxiii., p. 372. But no second examination of the fragment has been made, and it is not probable that the English authorities mistook reptilian for bird bones. Also no other reptilian bones have been found in these septaria, nor are reptilian bones known to occur in any other part of New Zealand associated with the species of shells found at Hampden in the beds containing the septaria: consequently, whatever age these beds may be, Mr. Forbes cannot be correct in saying that reptilian bones have been found on the same horizon; for geological horizons are determined by a similarity of fossils. Omitting this fragment as uncertain, we have undoubted traces of the moas in New Zealand in early Pliocene, or perhaps in Upper Miocene, times. Now comes the question, How came the moas to be in New Zealand? Mr. A. R. Wallace supposes that the moas are descended from Ratite ancestors of the Northern Hemisphere, and have spread southwards through New Guinea into Australia and New Zealand. If this be so the migration into New Zealand must have taken place a very long time ago. In the Miocene and Oligocene periods New Zealand was of much smaller dimensions than at present; but in the Eocene, and again, probably, in the older Pliocene, it was much larger, and stretched in a north-west direction towards New Guinea. But, as I have pointed out in former addresses to this Institute,‡ Ann. and Mag. Nat. Hist., ser. 4, vol. xiii., p. 25; ser. 5, vol. xiii., p. 425; and vol. xv., p. 77. N.Z. Journal of Science, 1884–85, pp. 1 and 249. biological evidence assures us that New Zealand has never been joined to Australia or to New Guinea since the Cretaceous period, a time when there were no struthious birds in existence; and it is difficult to explain how these birds, if they migrated from the Northern Hemisphere, entered Australia and New Zealand without being accompanied by placental mammals. Mr. Wallace supposes that they either flew or swam across a strait which was impassable to the mammalia.§ “Island Life,” p. 450. See also Owen, Ext. Birds of N.Z., p. 136. 10 That they flew across is, on this hypothesis, impossible, because the special characters of the Ratitæ are

due to their being unable to fly; and the only alternative is that they crossed by swimming. But, although the emu and the rhea are both said to take readily to water, many placental mammals do the same, and it is very unlikely that the struthious birds should twice have swum across straits—once from the Oriental to the Australian region, and again from the Australian region to New Zealand—which were impassable to mammals. There are also other reasons for doubting the northern origin of the Australasian Ratitæ. The struthious birds of New Zealand, notwithstanding the almost complete absence of wings, make a nearer approach to the original stock—that is, are less modified—than any of the other families. This is shown by the long lateral processes of the sternum, by the smaller coraco-scapular angle in the smaller species, by the broad sacrum and free ischia and pubes, as well as by the hind-toe. Again, the only small Ratitæ are from New Zealand; and, if the group is descended from flying birds, the smaller forms must have preceded the larger ones. We should expect to find the least altered forms near the place of origin; and, if the New Zealand Ratitæ have migrated from Europe or Asia, how comes it that the least-modified forms have been gathered together in these Islands? Flightless birds are generally developed on islands where there are no mammals, and not on large continents. The common ancestors of the Australian and New Zealand Ratitæ must have had both wings and hind-toes, like the kiwi and Palapteryx. Probably they also had, like the moas, feathers sometimes with one, sometimes with two, shafts to a quill. Palapteryx makes the nearest approach to the common ancestor of the Australasian Ratitæ of any bird we know; but no doubt these common ancestors had much larger wings than Palapteryx, and they must have possessed clavicles. The late Professor W. K. Parker long ago pointed out the struthious affinities of the tinamous,* Trans. Zool. Soc., vol. v., p. 149 (1866). and lately he has proposed to place the Ratitæ, and perhaps Opisthocomus, with the Tinamidæ in the group Dromæomorphæ of Huxley.† Trans. Zool. Soc., vol. xiii., p. 80 (1891). The tinamous are birds about the size of kiwis, living in Central and South America, and they show in the free ischia and pubes, as well as in having a free post-axial tarsal bone, a decided approach to the Dinornithidæ. Probably, therefore, the Australasian Ratitæ are descended from birds allied to the tinamous, which, being carinate and with large wings, found their way to New Zealand by flight, like all the other birds. There are no great difficulties in accounting for a migration by land of these birds from New Zealand into Australia, while

a counter-migration of Australian mammals into New Zealand was prevented. At some former period, when New Zealand stretched towards North Australia, a portion of land—including, perhaps, Norfolk Island—inhabited by Ratitæ may have been detached from New Zealand, and subsequently united to North Australia, when the Ratitæ would spread into Australia and New Guinea. If this supposition is a correct one the great differentiation that has taken place in the birds since their migration shows that it must have been not later than the Eocene period. According to Mr. De Vis a femur belonging to one of the Dinornithidæ has been lately found on the Darling Downs, in Queensland;* N.Z. Journal of Science, 1891, p. 97; and Etheridge, Records of Geol. Surv., N.S. Wales, vol. i., p. 128. and if this is correct it would imply that there had also been a second migration into Australia from New Zealand in the older Pliocene. But I cannot agree with Mr. De Vis that Dromornis was more closely related to the moas than to the emu. This question has been settled by Sir R. Owen from an examination not only of the femur, but also of the tibia and pelvis. The points of resemblance between Dromornis and Dinornis are adaptive only, and cannot be taken to prove affinity. If the absence of a pneumatic foramen in the femur shows relationship more strongly than other points, we should have to class the owls with the kiwis instead of with the hawks. We may conclude, therefore, that the ancestors of the moas originated in New Zealand in the Eocene period, although we have as yet no certain evidence of them before the older Pliocene or Upper Miocene. Probably the Ratitæ of America, Asia, and Africa, including Æpyornis, have had a separate origin from those of Australasia. The ancestors of Æpyornis must have passed into Madagascar in the Eocene period, for no Miocene mammals are found there, and it is possible that they may be descended from swimming-birds through Gastornis, Dasornis, Megalornis, and Diatryma. Development of the Moas in New Zealand. In whatever way the moas originated in New Zealand, it is evident that the land was a favourable one, for they multiplied enormously, and spread from one end to the other, no doubt during a time when both Islands were connected. Bones have been found from the Bay of Islands in the north to Waipapa Point in Southland, and from sea-level to an altitude of 5,250ft. in the mountains of Otago.† N.Z. Journal of Science, 1885, p. 394. The moa-bones found on the Great Barrier Island by Mr. Weetman (Trans. N.Z. Inst., vol xix., p. 193, and vol. xxii., p. 84) belong to a young chick of a species of Dinornis, and possibly the bird was taken there by the Maoris. No moa-bones have been found on the Chatham Islands by Europeans, but the Maorioris have a tradition of a large bird which was called puoa (Trans. N.Z. Inst., vol. vii., p. 117, footnote). The number of individuals

living together must have been very great, if we may judge by the number of bones found in the swamps and in the alluvial deposits of rivers. Writing in 1871, Sir James Hector says, “It is impossible to convey an idea of the profusion of bones which, only a few years ago, were found in this district [central Otago], scattered on the surface of the ground, or buried in the alluvial soil in the neighbourhood of streams and rivers.”* Trans. N.Z. Inst., vol. iv., p. 115. Sir Julius von Haast estimated that there were the remains of more than a thousand birds, belonging to fourteen different species, in the Glenmark Swamp. At Hamilton I obtained from a small basin, about 50ft. in diameter and 4ft. in depth, bones of at least four hundred birds, belonging to eleven species; and Mr. Forbes estimates that at Enfield the remains of more than three hundred birds were crowded together into a small space not more than 3ft. deep. The great number of different species of struthious birds lately living together in New Zealand is a remarkable fact, unparalleled in any other part of the world.† Owen, Ext. Birds of N.Z., p. 106. The Continent of Africa, together with Arabia and, formerly, central Asia, contains but three species of ostrich, differing in the colour of the skin of the neck. South America, from the Straits of Magellan to Peru, has but three species of rhea. Australia possesses two species of emu and one species of cassowary; while eight other species of cassowary inhabit detached islands from New Britain and New Guinea to the Aru Islands and Ceram. Outside of New Zealand two species of struthious birds are hardly ever found living in the same district; while a few hundred years ago there were in New Zealand, besides several kinds of kiwis, twelve species of moas in the North Island and seventeen species in the South Island. Many years ago I found, as I thought, a solution for this problem by examining the present distribution of the cassowaries.‡ Trans. N.Z. Inst., vol. v., p. 233. Here we have eight species, inhabiting five different islands, and if this region of the earth were to be elevated and the islands joined these eight species might mingle together. If the region were to sink again all would undoubtedly be driven to the highest land, and we might have a single island inhabited by eight species of cassowaries. Now, we know that New Zealand has actually gone through a series of changes in level similar to those just mentioned. In the Oligocene and Miocene periods it consisted of a cluster of several islands,

which were united in the older Pliocene, and ultimately it divided into the two islands which we have now. If the ancestors of the moas inhabited New Zealand during the Eocene period—as we have seen was most probably the case—each island in the Miocene may have contained a different kind of moa, all of which would mingle together when the land rose in early Pliocene times. In the newer Pliocene the birds must have been again separated by the formation of Cook Strait, and the species in each island would again be isolated from each other. I still think that isolation during Miocene times gave origin to the genera; but variations of specific importance have in several cases taken place since the formation of Cook Strait, and most of the species must be due solely to variation without isolation. As is the case with most common species, the moas varied greatly, and, there being no carnivorous mammals or other powerful enemy to hold them in check, while vegetable food abounded in all directions, natural selection did not come into play, the intermediate forms were not strictly eliminated, and, consequently, the different species were not distinctly marked off, but one merged into the other. Under such favourable circumstances the conditions of life were very easy, and the birds got larger and fatter, more sluggish and more stupid. If the moas are descended from flying birds, which we cannot doubt, it is evident that the smaller species must have preceded the large ones; and this is borne out by the facts of geographical distribution. For it is only the smaller species of Anomalopteryx, Mesopteryx, Syornis, and Dinornis which are found in both Islands, the larger forms in Syornis and in Dinornis having been developed since the Islands were separated. Whether the apparently allied species of Dinornis are really geographical representatives of each other, due to isolation, or whether they form two separate lines, one in each Island, diverging from D. struthioides, we cannot say until the skulls and sterna of the North Island species are known; but the latter hypothesis seems to me to be the more probable, as it accounts for all the species of the South Island being more robust than those from the North Island. When we try by morphological evidence to reconstruct the genealogy of the moas we see that Mesopteryx connects Cela with Syornis, that it is itself connected with Euryapteryx by Anomalopteryx, and that this last genus is connected with Dinornis by Palapteryx. Evidently Anomalopteryx and Palapteryx are the oldest forms: but, if Palapteryx had wings, it could not have been derived from the wingless Anomalopteryx; and, if the birds were increasing in size, Anomalopteryx could not have been derived from Palapteryx. Both must have had a common

ancestor as yet unknown, and we thus get the following classificatory diagram:— Palæontology throws very little light on the subject. All the species—except A. antiquus, which is Pliocene only—have been found in the Pleistocene peat deposits, and all of them except six—D. altus, D. maximus, D. excelsus, D. validus, D. torosus, and P. plenus—have been found in recent Maori cooking-places. The following species from the older Pleistocene beds at Motunau are represented in the Wellington Museum: D. potens, D. torosus, S. rheides, S. crassus, S. casuarinus, E. elephantopus, E. ponderosus, and E. gravis; and the bones of the three species of Euryapteryx have the proportions of the most robust specimens found in younger deposits. Of the Pliocene species we know but little. Professor Owen has referred the footprints sent to him, in the sandstone from Poverty Bay, to D. ingens, D. struthioides, and P. dromioides. Those described by Archdeacon Williams and Mr. Justice Gillies were made by a bird not larger than A. didiformis, which had a foot about 8in. in length. The bones of A. antiquus found by Mr. Stubbs under the lava-stream at Timaru are the oldest known moa-bones, and they belong to a species which was smaller than any which lived subsequently in the South Island.* Mr. Forbes is mistaken in saying of these bones that “the largest, nearly 8in. in length, were undoubtedly portions of Dinornis bones of one of the greater forms” (Trans. N.Z. Inst., vol. xxiii., p. 367). I have examined Mr. Stubbs's collection, and there is in it no evidence of a moa larger than A. antiquus. It seems, therefore, that very little change took place in the moas subsequent to the date of the oldest beds at Glenmark and Motunau. No new species, or even varieties, were formed, but possibly some of the larger forms of Dinornis may have died out. But even this is not certain, for the large species of Dinornis are rare everywhere, and it is possible that they may have lived away from the coast, in which case we can easily explain the absence of their bones in the Maori cooking-places, for, except in central Otago, it is only the sand-dunes along

the coast which have preserved to us the relics of the ancient moa-hunters. But considerable change must have taken place in the moas between the time of the formation of Cook Strait and the deposition of the Motunau beds, because out of the eight species obtained from there only one is found in the North Island, and out of the whole twenty-five Pleistocene species only four are common to both Islands. We must therefore suppose that the other twenty-one species are of later origin. If, then, the Motunau beds are Pleistocene, we must put the formation of Cook Strait back into the Pliocene. It is also evident that the genera Dinornis, Tylopteryx, Palapteryx, Anomalopteryx, Mesopteryx, and Syornis had all been differentiated before the formation of Cook Strait, for their remains are found in both Islands, while the genera Cela and Euryapteryx may have come into existence at a later date. This great differentiation implies a long interval of time, and it probably took place during the isolation of the species in the Oligocene and Miocene periods, so that we again arrive at the conclusion that the ancestors of the moas inhabited New Zealand during the Eocene period. It should be remembered that the Eocene includes a duration of time probably equal to the Oligocene, Miocene, and Pliocene together. In the Pliocene period the moas must have flourished greatly and covered the land. Euryapteryx had its head-quarters in Otago, Syornis in Canterbury, Mesopteryx in both Otago and Canterbury. Anomalopteryx and Palapteryx were most common in Nelson and Wellington, Dinornis in Hawke's Bay, and Cela in Hawke's Bay and in Auckland. But it would seem that in the Pleistocene period the moas suffered great mortality in the South Island, for how else could such great quantities of bones, of both young and old birds, have got together in the swamps at Hamilton and Glenmark? It has often been suggested that flocks of the birds, attempting to escape from fires, rushed into the swamps and perished; but when we remember that these moas died thousands of years ago, long before there were any human inhabitants to light the fires, it will be seen that this surmise is quite out of the question. Only two hypotheses, or a combination of them, appear possible to account for the facts. Either the birds walked into the swamp and there perished, or their dead bodies were washed in. If we suppose that a swamp acted as a trap for birds trying to cross it, we can conceive that in time a great number of birds may have been trapped in the same hole. According to Maori tradition the moas frequented wet places; and the fact that these swamps always contain the remains of a large number of young birds is also favourable to the idea. This theory, originated by Mr. Mantell, is thought by both Mr. A. Hamilton and Mr. Park to

account for the bones in the Te Aute Swamp, because most of the larger beg-bones were found in a vertical position, the tibia and metatarsus often in their relative places; and the same, to some extent, was the case at Waikouaiti.* Mantell, “Petrifactions and their Teachings,” p. 98. But the position of the bones at Glenmark and Hamilton was very different, the leg-bones lying in all directions, and just as often upside down as in any other position.† Booth, Trans. N.Z. Inst., vol. vii., p. 128. It is also difficult to account by this theory for the swamping of tuataras, Cnemiornis, Harpagornis, the kiwi, and still smaller flying birds, which have been found in the swamps with the moa-bones. Sir J. von Haast thought that the first hypothesis might account for the occurrence of bones in the main swamp at Glenmark; but the older peat-beds, like those at Hamilton, are not deep enough to swamp a moa, and the bones go to the very top of the bed. At Hamilton, and also at Glenmark, there is a considerable amount of evidence in favour of the second hypothesis, that the bones were washed in by floods; but, as none of the bones are waterworn, and the peat always contains a large number of moa-stones, whole birds in the flesh must have been washed in, and not single bones. At Glenmark, wherever a small watercourse came in from the surrounding hills, a network of drift timber, often of large size, with numerous moa-bones, was always found round its mouth. In the swamp the bones occurred in “nests” near clay banks, while in other places there were no bones. Dr. Haast remarks that “the carcases or portions of birds had evidently been washed here against the banks, and deposited in considerable quantities.”‡ “Geology of Canterbury and Westland,” p. 446. At Hamilton very little timber was found, probably because it had all decayed into peat, but stones up to 1lb. and 2lb. occurred, and one piece of rock weighing between 10lb. and 12lb. was found in the clay near the bones. As the heaviest moastone does not weigh much more than 2oz., these large stones must have been entangled in tree-roots which had been washed into the swamp by floods.§ Quar. Jour. Geol. Soc., vol. xli., p. 213. The peat-bed at Hamilton contained numbers of fresh-water shells, and it probably formed the lowest portion of a former lake, into which the materials brought down by floods from the surrounding hills collected. We find corroborative evidence of this in the alluvial or old lacustrine deposits all round the plains of central Otago, for these always contain numerous bones wherever a stream enters them from the hills. Mr. Vincent Pyke—a very early

colonist, and acute observer—says that at Moa Flat, in Otago, the bones were scattered around the plains on the old lake-terraces, or in shallow ravines washed out by the floods and rain. “Around these ancient lake-beds the remains of the moa were most commonly found in heaps, in piles, in layers, and so found in a state of what may fairly be termed perfect preservation, with moa-stones underneath each skeleton…. they were simply lying on the surface.” He also “observed the same thing on the terraces, and in shallow gullies evidently washed out by rain, but always at a certain elevation—at Hamilton and other parts of the Maniototo Plains—but never on the plains.”* “The Moa,” Wellington, 1890. Sir James Hector also says, “The greatest number of moa-bones were found where the rivers debouch on the plains.”† Trans. N.Z. Inst., vol. iv., p. 115. Mr. Booth says, “I find below a certain level that would leave the whole Maniototo Plain under water there are no moa-bones to be found with the exception of near the mouths of the burns coming in from the hill, where the bones have been brought down by freshets…. I have crossed these plains in several places, and do not recollect ever having seen a moa-bone below the level spoken of. I have also inquired of several shepherds who live on the plain, and who have walked over every acre of it, and none of them recollect having seen a moa-bone below the level I have mentioned.”‡ Trans. N.Z. Inst., vol. vii., pp. 132, 133. This evidence seems to be sufficient. But how are we to account for the number of dead birds washed down by the floods? It is a remarkable fact that, while the bones of young moas are rare in Maori cooking-places and in caves, they formed a very large proportion of the bones at Glenmark and at Hamilton: indeed, in the older peat-bed at Hamilton, under the clay, the bones of young birds were as numerous as those of adults. Most of these young birds are from one-half to three-quarters grown, mixed with a few others. Another interesting fact is that neither at Hamilton nor at Glenmark was a single fragment of moa-eggshell found.§ A piece of eggshell is reported as having been found at Enfield. We cannot suppose that all the eggshell has been dissolved, because at Hamilton I collected from the peat many delicate land and fresh-water shells, and in the swamps of Madagascar fragments of eggshell of Æpyornis are said to be not uncommon. We must conclude that the female birds died at a time of the year when they contained no hardened eggs; and this, together with the large number of young birds, points to the autumn or winter as the time when the moas died. It is also evident

that the dead moas could not be washed into swamps under the present climatic conditions, and the solution of the problem is to be found in the fact that in Pleistocene times, when these deposits of bones were formed, the climate was very different from what it is now. At that time the eccentricity of the earth's orbit was very great, and, when winter in the Southern Hemisphere happened in aphelion, long cold winters were followed by short and very hot summers, which would produce what has been called a pluvial, or, better, diluvial period.* For proofs of a pluvial Pleistocene period in Australia see Wilkinson, Proc. Linn. Soc. N.S. Wales, vol. ix., p. 1227 (1884). For New Zealand, see Quar. Jour. Geol. Soc., vol. xli., p. 213 (1885). The snows of the early winters would kill large numbers of moas and other birds on the hills, and the summer floods and avalanches would deposit their bodies in hollows, or on the low ground at the foot of the hills. As the Pleistocene period passed away the climate no doubt got more equable, and the surviving moas once more increased and multiplied. Some, perhaps, of the larger forms had succumbed altogether before human beings visited the islands, but most of the species were still living at that time, and were subsequently exterminated by the hand of man. Extinction of the Moas. The first collection of moa-bones, made by Mr. Colenso, Mr. Williams, and Mr. Cotton, was obtained by Maoris chiefly from the river-beds in the Waiapu and Poverty Bay districts. In 1842 Mr. Colenso said that, although the true age of the bones was not certainly known, he thought that “they will be found lying in the upper stratum of the Secondary [Tertiary] or the lower strata of the Tertiary [alluvial] formation,” and probably “in beds of shingle, the detritus of the deluge.”† Tasmanian Jour. Science, 1843, vol. ii., p. 87; and Trans. N.Z. Inst., vol. xii., p. 72. “From native tradition,” he said, “we gain nothing to aid us in our inquiries after the probable age in which the animals lived; for, although the New-Zealander abounds in traditionary lore, both natural and supernatural, he appears to be totally ignorant of anything concerning the moa save the fabulous stories already referred to.” The collections also of Mr. Percy Earl and Dr. Mackellar, in 1846, were from an old turbary deposit at Island Point, Waikouaiti, of Pleistocene age.‡ Mantell, Quar. Jour. Geol. Soc., vol. iv., p. 238; and Amer. Assoc. Proc., 1850, p. 252. In 1847 the Hon. W. Mantell collected from the old Maori cooking-places at Waingongoro, near Wanganui, and he was

the first to recognise that the moa had been killed and eaten by man. In 1852 he obtained more proofs of this at Awamoa, near Oamaru,* Trans. N.Z. Inst., vol. xxi., p. 440. and he further ascertained that the Maoris had “traditions concerning the existence of the moa and the use of it by them for food, of its bones for implements, and its feathers for ornaments.”† Trans. N.Z. Inst., vol. v., p. 95. In 1851 he read a paper on the subject to the New Zealand Society which, I believe, was never published; and in the discussion that followed the reading to the Zoological Society of London, in 1856, of Professor Owen's paper on the Awamoa Collection, Mr. Mantell, who was present, successfully combated the then prevailing idea in England that the moa was Pleistocene only. In 1864 Sir W. Buller published a letter in the Zoologist, in which he said that the moa is now extinct, but was contemporaneous with the Maoris, as is proved by the broken and calcined bones in the refuse of their feasts, and by “the rude history of the bird preserved in Maori tradition.” In the same year the skeleton of D. potens, which is now in the York Museum, was found at Tiger Hill; and, as this skeleton had portions of the skin and ligaments attached, it completed the reversal of opinion among English naturalists. In June of that year Mr. Allis read a paper about it to the Linnean Society, when the general opinion of the meeting was that probably the bird had been living within ten years;‡ Zoologist, 1864, p. 9195. and in 1868 Mr. E. Newman concluded that the last moa died in about the year 1800 or, perhaps, later.§ Zoologist, 1868, p. 1354. But in the same year the Hon. W. Mantell, in a lecture to the New Zealand Institute, pointed out that the extermination of the moa must have taken place within a very short period after the appearance of man on these Islands, as the allusions to the bird in the most ancient Maori traditions are very slight and obscure. He also said that nephrite appears to have been discovered at a later date than the extinction of the moas, because it was never found in the Maori cooking-places with moa-bones.∥ Trans. N.Z. Inst., vol. i., p. 18. Three years later Sir J. von Haast published his papers on “Moas and Moa-hunters,”¶ Trans. N.Z. Inst., vol. iv., p. 66. in which he denied the existence of any Maori traditions about the moa, and held that the birds had been exterminated by men “most probably belonging to a different race from the present native inhabitants of the Islands” (p. 68)—a race to whom not only was greenstone unknown, but who had not even acquired the art of grinding

stone adzes, and who had no domesticated dog—a race which had passed away “long before the Maoris settled here” (p. 68). In fact, he said “he might even assume that the human race [i.e., the moa-hunters] made its appearance when this [land] eommunication existed” between the two Islands (p. 84); for, he said, so rude a people could hardly have built eanoes. But a little further on he thinks that possibly the moa-hunters were identical with a race which, according to Mr. A. Mackay, formerly existed in the interior forests of the North Island, and were called “Maeros” by the Maoris; and in his “Geology of Canterbury and Westland” (1879) he calls them “an autochthonie race having affinities with the Melanesian type” (p. 430). Six months after the reading of Dr. Haast's first paper Mr. Murison stated that he had found in central Otago polished stone implements with moa-bones;* Trans. N.Z. Inst., vol. iv., p. 122. and in 1874 Dr. Haast himself found ground stone implements with moa-bones in a cave near Sumner. Consequently he withdrew his former opinion on this point,† Trans. N.Z. Inst., vol. vii., p. 72. and said that the moa-hunters “had reached already a certain stage of civilisation, which in many respects seems to have been not inferior to that possessed by the Maoris when New Zealand was first inhabited by Europeans” (I.e., p. 80). The only other reasons advanced by Dr. Haast for disassociating the moa-hunters from the Maoris were that the moa-hunters were not cannibals, and that they did not possess a domesticated dog. The first of these reasons, although correct for the South Island, is not correct for the North Island. It was opposed to the evidence of Mr. Mantell,‡ Trans. N.Z. Inst., vol. i., p. 18. and was subsequently disproved by Mr. Thorne; § Trans. N.Z. Inst., vol. viii., p. 88. while the negative evidence on which the absence of a domesticated dog was inferred was disposed of by Mr. Booth, who found two moa-bones marked by the teeth of dogs in the old Maori cooking-places at the mouth of the Shag River. ∥ parallel Trans. Inst., vol. viii., p. 106. In 1883 M. de Quatrefages summed up the published evidence, and came to the conclusion that the moa had probably lived to about the year 1770 or 1780. ¶ Ann. and Mag. of Nat. History, ser. 5, vol. xiv., pp. 124 and 159. So far there had been no distinct proof of nephrite having been used by the moa-hunters;** Sir James Hector says that Mr. Murison “found polished adzes of apbanite, and even jade,” in the cooking-places of the Maoris (Trans. N.Z. Inst., vol. iv., p. 117); but Mr. Murison makes no mention in his own paper of jade implements having been found. but in 1889 Monck's Cave was discovered, near

Sumner, in which polished nephrite and other stone implements, as well as wood carvings of Maori pattern, were found, with moa-bones, and fragments of moa-eggshell still retaining the shell-membrane.* Trans. N.Z. Inst., vol. xxii., p. 64, and vol. xxiii., p. 373. There can therefore be no longer any doubt that the moas were exterminated by men of the Maori race; and the only question remaining is, How long was that ago? In the North Island we have to trust almost entirely to traditional evidence. Mr. J. S. Polack, in his book, “New Zealand,” published in 1838, says that “the natives [of the East Cape district] added that in times long past they received the tradition that very large birds had existed [in New Zealand]; but the scarcity of animal food, as well as the easy method of entrapping them, had caused their extermination.”† Vol. i., p. 303. Quoted by Hector, Trans. N.Z. Inst., vol. v., p. 413, footnote. He adds, “I feel assured, from many reports I received from the natives, that a species of Struthio still exists in that interesting [South] Island, in parts which perhaps have never yet been trodden by man. Traditions are current among the elder natives of atuas, covered with hair, in the form of birds, having waylaid former native travellers among the forest wilds, vanquishing them with an overpowering strength, killing and devouring, &c.” Mr. Polack gives no name to these birds, but Mr. W. Colenso says that in 1838 the Maoris of the same district had fabulous traditions of a large bird, which they called moa.‡ Trans. N.Z. Inst., vol. xii., p. 64. “Moa” is the native name for the domestic fowl in Polynesia; but, as the Maoris do not appear to have brought the fowl with them to New Zealand—along with the dog, the kumara (Convolvulus batata), and the taro (Colocasia antiquorum)—it seems possible that it was not known in Polynesia at the date of the Maori migration. If this conjecture be correct, the word” moa” must have been used for some other bird, and it has been suggested that it meant the cassowary. This, however, is hardly, possible, because the cassowary is not found on any of the islands from whence the Maoris are supposed to have come, but is confined to Melanesia and the Molucca Islands. It is called “mooruk “in New Britain. Mr. Rule, who brought the first bone to Professor Owen in 1839, told him that the Maoris had a tradition that this bone belonged to an extinct hawk, which they called “movie.”§ The extinct eagle, Harpagornis, appears to be known in Maori tradition as the hokioi (see Trans. N.Z. Inst., vol. v., p. 435, and vol. xii., p. 99). In the South Island the same bird, or an ally, was called “pouakai” (I.e., vol. x., p. 63). In the same year the Rev. W. Williams and the Rev. R. Taylor found bones of a large bird near Waiapu, which Mr. Williams says the Maoris called “moa,” while Mr. Taylor says that they called

it “tarepo.”* Trans. N.Z. Inst., vol. v., p. 97. In a letter to Sir E. Home in 1844 he says, “kakapo or tarepo” (Trans. Zool. Soc., vol. iii., p. 32). Mr. W. Travers says that “tarepo” was the native name for the extinct goose Cnemiornis (Trans. N.Z. Inst., vol. viii., p. 75, footnote). Three years later (1842) he says that the Maoris of Waingongoro, near Wanganui, called the bird “moa.” Mr. J. W. Hamilton says that in 1844 the Europeans knew very little about the existence of moa-bones, and very few had at that time been found; but the Maoris always knew them when they saw them.† Trans. N.Z. Inst., vol. vii., p. 122. It is evident, therefore, that the Maoris had a tradition that these bones, which they used for fish-hooks, had belonged to birds which they called “moas.” Mr. W. Colenso, in a very valuable and thoroughly scientific paper on the subject,‡ Trans. N.Z. Inst., vol. xii., p. 63. distinctly recognises that the ancestors of the Maoris knew the moa, but says that this knowledge dates from “very long ago, in almost prehistoric times, long before the beginning of the genealogical descent of the tribes, which, as we know, extended back for more than twenty-five generations; “for the moa is rarely mentioned in their poetry or proverbs, and even then the allusions are largely mythical. But the name “moa,” he says, is incorporated in many words handed down from early ancestors— firstly, as names of places, such as Moawhiti (startled moa), Moakura (red or brownish .moa), Moarahi (big moa), Otamoa (moa eaten raw), Moawhango (hoarse-sounding moa), and others: secondly, in names of persons, as Hinemoa, “hine” meaning daughter of rank, or young lady: and, thirdly, as ordinary words—for example, Maimoa (Come-hither moa), used as the name for a decoy-bird; Moamoa (small round stones the size of marbles), perhaps a reminiscence of moa gizzard-stones. There is also a tradition among the Maoris of the East Cape district that the moas were exterminated by a fire known as the fire of Tamatea; and Mr. Colenso remarks that Tamatea is a very ancient name in New Zealand mythological history. He was a descendant of Toto, and, with his children, came to New Zealand in the canoe Takitumu. Mr. Mantell says that “the extermination of the moas must have taken place within a very short period after the Maoris reached the Islands, as the allusions to the birds in their most ancient traditions are very slight and obscure.”§ Trans. N.Z. Inst., vol. i., p. 18. Major W. G. Mair, who takes great interest in science, has a thorough knowledge of the Maori language, and has been for many years collecting Maori tales, says,” In all these thousands of pages of Maori lore which I have written from the mouths of

[Maori] witnesses in Waikato, at Rotorua, in the Bay of Plenty, Hawke's Bay, Manawatu, Wanganui, and Taupo, there is not one word about the moa.”* Trans. N.Z. Inst., vol. xxii., p. 74. The Rev. J. W. Stack says the same; and he has pointed out that the saying, “Ka ngaro i te ngaro a te moa” (“Lost as the moa is lost”), or, as Mr. Colenso translates it, “All have perished just as the moas have perished,” which occurs in the very ancient Maori poem called the “Lament of Ikaherengatu,” shows that the moa was not in existence at the time when it was composed.† Trans. N.Z. Inst., vol. vii., App., p. xxviii. On the other hand, the Rev. R. Taylor,‡ Trans. N.Z. Inst., vol. v., p. 100. Judge Maning,§ Trans. N.Z. Inst., vol. viii., p. 102. Mr. John White, ∥ parallel Trans. N.Z. Inst., vol. viii., p. 79. Sir George Grey,¶ Proc. Zool. Soc., 1870; Zoologist, 1870, p. 2104. Sir W. Buller,** Birds of N.Z., 2nd ed., vol. i., p. xxxiii. Lieut.-Colonel McDonnell, †† Trans. N.Z. Inst., vol. xxi., p. 438. and Judge Monro‡‡ Trans. N.Z. Inst., vol. viii., p. 427. all agree that the present Maoris have plenty of traditions about the moas. Sir G. Grey says that when he came to New Zealand, in 1845, the Maoris invariably spoke to him of the moa as a bird well known to their ancestors; and Sir Walter Buller says that their “ancient folk-lore, their historical songs, and their proverbial sayings are full of allusions, more or less direct, to the bird.” Colonel McDonnell was told that the moa was of a brown colour, with feathers longer and coarser than those of the kiwi; that it fought fiercely when brought to bay, and struck out with its feet, but was easily killed with clubs. §§ None of the skulls found in Maori cooking-places have been broken with clubs. Kawana Paipai's statement that he had himself hunted the moa on the Waimate Plains, Taranaki, is no doubt a romance. According to Judge Maning they were stupid and sluggish birds, and were killed in great numbers by fires. Periodically they fought with great fierceness. According to Mr. John White the moas did not go in flocks, but lived in pairs with their young. The same writer says that the nest was merely a heap of grass, on the top of which they laid their eggs; while the Rev. R. Taylor says that the nests were made of refuse fern-root on which they fed. Mr. White says that they lived principally on the young shoots and roots of fern and grass, as well as on the shoots of a shrub with yellow flowers, called korokia (Corokia buddleoides), which grows on the margins of the bush. They haunted chiefly the edges of the forest, but often visited lakes and water-pools to feed on water-plants. The Maoris used to lie in wait for them near the tracks by which they went to the water, and strike them

with spears, which were so cut that they broke off and left about 6in. or 8in. of spear in the bird. The bird were driven from the water by one party of men, while another party lay in ambush to spear them. Mr. Taylor gives a slightly different account of a moahunt. He says that notice was given to neighbouring settlements that a hunt was to take place. A large party spread out to drive the birds from their haunts towards a lake or swamp. As they approached the water the line of beaters was gradually contracted, until at last a rush was made with loud yells, and the frightened birds were driven into the water, where they could be easily approached in canoes and despatched without their being able to make any resistance. According to Sir G. Grey, Mr. Stack, and Mr. White, the Maoris always used koromiko (Veronica salicifolia) for cooking the flesh. They have a saying, “The koromiko is the tree which roasted the moa.” But Mr. Colenso has pointed out that the wood of the koromiko is too small to heat a Maori oven, and he suggests that the saying should perhaps be translated “which burnt the moa.” At first sight all this seems to be in direct conflict with the opinions of Mr. Colenso, Mr. Mantell, and Major Mair; but we must remember that none of these traditions have been published in Maori, and many of them may be nothing but late deductions from the words and proverbial sayings mentioned by Mr. Colenso, for it is evident that they were not generally known among the Maoris. At any rate, it is quite time that the evidence for these being ancient Maori traditions should be put upon record. But Sir Walter Buller has given a story of the pet moa of the Ngatituwharetoa which cannot have been so derived. He says that this bird was stolen by a man named Apa-hapai-taketake, an ancestor of the Ngatiapa Tribe. A series of fights ensued, in which the Ngatiapa came off worst, and were driven south from the Bay of Plenty to Lake Taupo.* Birds of N., 2nd ed., vol. i., p. xxviii. I have not been able to ascertain the probable time of this migration, but I judge from the context that the Ngatiapa took up unoccupied ground near Taupo, and if this be so Apa-hapai-taketake must have been a very early ancestor. At the present time the tribe lives on the West Coast, between Wanganui and Otaki. In Mr. White's large work on the “Ancient History of the Maori,” I can only find two allusions to the moa. One is about its extinction by fire in the South Island, which is copied from Mr. Stack; the other relates to its first discovery. “Nga-hue was the first man who came to these Islands, who at Te-wai-rere saw the bird moa, and

killed one, and went back to Hawa-iki and told the inhabitants of that land that he had discovered a country without human inhabitants, but where there was greenstone to be found.”* Anc. Hist. of the Maori, vol. ii., p. 187. A little further on the moa is said to have been killed near the Wai-rere waterfall at Arahura [nearHokitika], and that Ngahue carried it away in a calabash. Major Mair gives another story about a man named Hape, who pursued a moa to the top of a hill in the Rotorua district. When trying to catch it the moa struck out and hurled him back into the valley. But, being a “tangata atua” [god-man], he was not killed, and his heel, striking against a rock, split it, and caused the water of the Tarawera River to flow in its present subterranean channel at Te Tatau-a-Hape.† Trans. N.Z. Inst., vol. xxii., p. 72. The only other published story about the moa is the well-known one of the bird supposed to be living in a cave on Whakapunake Mountain, guarded by a dragon, but seen by none.‡ Trans. N.Z. Inst., vol. xii., p. 67, and vol. xxii., p. 72. 11 In discussions on this subject much has been said about positive and negative evidence. It has been even affirmed that “the testimony of a man who had actually seen and eaten the moa was worth ten thousand legends and traditions.” But there is here, I think, a misconception, for the relative value of the two kinds of evidence depends altogether on the application. We believe implicitly many things on negative evidence alone. For instance, we believe that there are no snakes in New Zealand entirely on negative evidence. We believe this because we feel sure that if there were snakes in New Zealand some would have been found before now. And, so far as the North Island is concerned, I am compelled to believe that the moas were exterminated many years ago, because I feel sure that if it were not so we should find as many allusions to it in Maori tales and poetry as we do to all the other birds, beasts, and fishes that were of interest to the natives. The question really hinges on the reliability of the evidence, and not whether it is negative or positive. In this case the negative evidence is the more reliable, for it is unbiassed, and existed before the question was raised. But the positive evidence is not all on one side. The very earliest statement of the Maoris to Mr. Polack was that “in times long past they received the tradition that very large birds had existed;” and the saying, “Ka ngaro i to ngaro a te moa,” is positive evidence that the Maoris who recited it believed the moa to have been lost or exterminated when this very ancient poem was composed. This appears to me to be conclusive proof that the moa was exterminated rapidly, soon after the arrival

of the Maoris in New Zealand. The dodo of Mauritius existed for seventy-three years only after the island was discovered; and the sea-cow of Behring Strait, which was living in immense numbers in 1741, succumbed entirely in twenty-seven years. The Maoris are supposed to have inhabited New Zealand for about six hundred years,*Writers are of opinion that the ancestral records of the Maoris of the North Island prove from eighteen (Shortland) or twenty-five (Colenso) to forty-six (J. White) generations since their first arrival. so that we must put the probable extinction of the moa in the North Island at four or five hundred years ago. In the South Island there are no Maori names of places containing the word moa;† Awamoa, near Oamaru, was so named by the Hon. W. Mantell. and Mr. A. Mackay, the Rev. J. F. H. Wohlers, and the Rev. J. W. Stack all agree that the natives have no traditions of the moas further than that they were destroyed by a fire, called the fire of Tamatea, which swept over the Canterbury Plains some five hundred years ago—evidently an echo of the legend of the North Island already mentioned. According to Mr. Stack tradition states that the first occupants of the South Island were the Kahui-tipua, a fabulous race of giants who ate men. They were destroyed by Te Rapu-wai, who were soon followed, from the North Island, by the Wai-taha. These two belonged probably to the same tribe; at any rate, they were contemporaries, intermarried freely, were not warlike, became numerous, and “covered the land like ants.” The Waitaha were exterminated by the Nga-ti-mamoe, who crossed Cook Strait into the South Island about three hundred years ago; and the Ngatimamoe were, in their turn, destroyed by the ancestors of the present natives—the Nga-i-tahu—about two hundred or two hundred and fifty years ago. Very little is known about Te Rapu-wai and Waitaha. Their traditions perished with the extinction of their conquerors, the Ngatimamoe, but the extinction of the moa, as well as the formation of the shell-heaps on the coast, are attributed sometimes to one, sometimes to the other.‡ Stack, Trans. N.Z. Inst., vol. x., p. 60,&c. Copied by Mr. White in his “Ancient History of the Maori,” vol. iii., p. 191. See, also, Mantell, Trans., Trans. N.Z. Inst., vol. xxi., p. 440. It is evident that Haumatangi, who told Mr. J. W. Hamilton that he had himself seen the last moa, and the Rapaki (Lyttelton) Maori who told him that his father had hunted the moa,§ Trans. N.Z. Inst., vol. vii., p. 121. were romancing. As also was the sealer Meurant, who said that he had seen and eaten moa's flesh at Port Molyneux as late as 1823. ∥ parallel Trans. N.Z. Inst., vol.vii., p. 121. We must also consider as fabulous the Maori statements made to:Mr.

W. H. Roberts that the moa was “a very swift bird on foot, and could outrun a horse easily, its miniature wings helping, it considerably;” and that they used to steal children.* Trans. N.Z. Inst., vol. vii. p. 548. Also the story of the man-eating bird of the forests of the South Island mentioned by Mr. Polack. No one knows the Maoris of the South Island better than the Rev. J. W. Stack and the Rev. J. F. Wohlers, and yet neither of them ever heard of these legends. Mr. Stack says, “The Ngaitahu have occupied this [South] Island for about ten generations. Allowing twenty-five years for a generation, their occupation dates back two hundred and fifty years. In none of the traditions relating to this period, though numerous and detailed, are there any allusions to the moa.”† Trans. N.Z. Inst., vol. iv., p. 107. Mr. Wohlers has published a number of southern Maori tales, ‡ Trans. N.Z. Inst., vol. vii., p. 3, and vol. viii., p. 108. collected more than forty years ago, “when there were still a few Maoris alive who were acquainted with their ancient lore;” and, although whales, seals, dogs, rats, owls, tuis, pigeons, herons, eels, and other animals are mentioned, the word “moa” does not once occur. However, we find here the legend of the man-eating bird. Two strangers arrived from Hawaiki and taught the natives how to make fire and cook their food. The natives told the strangers that they were suffering from a monstrous bird which ate people. “The strangers asked if the direction the bird used to take was known. ‘Yes,’ was the reply, ‘a nd, if some of us go that way when the bird happens to come, he gobbles us up.’ Having learned this they went to the haunts of the bird and erected a sort of blockhouse, having only one small opening at some height from the ground, into which they jumped, and then waited for the appearance of the bird. After some time they saw it coming. The body was still at a distance when the head reached their little fortress. The bird came nearer and raised its huge beak towards the opening where the men stood, but the throw of a heavy axe from the men broke one of its wings. Again it raised its beak, and again an axe broke its other wing. Then the men jumped down and killed it. After that they went to its cave, and found there a heap of human bones.”§ Trans. N.Z. Inst., vol. viii., p. 109. Every one will, I think, allow that this very ancient legend was not recited or even composed by men who had a personal knowledge of the moa. There is also a story of catching a bird by putting a rope over the entrance to the cave in which the bird was living.∥ parallel L.c., vol. vii., p. 22. This may be the origin of Whera's

statement to Mr. J. W. Hamilton about catching moas with a rope;* Trans. N.Z. Inst., vol. vii., p. 121. but in the ancient tale the bird is said to have been the blue heron, and not the moa. Lastly, we must consider the evidence afforded by the finding of nephrite with moa-hunter remains in Monck's Cave, near Sumner. The Ngaitahu have a tradition that their ancestors knew nothing of nephrite until a piece was brought over from Hokitika, in about the year 1700, by a woman named Raureka.† Stack, Trans. N.Z. Inst., vol. x., p. 86. But, allowing this to be correct, it does not follow that Te Rapu-wai and Waitaha were ignorant of it. The knowledge of the places where nephrite was found might easily have perished with them, or with the Ngatimamoe; and another tradition that the Ngaitahu came to the South Island in order to find nephrite is quite in accordance with this idea. Indeed, tradition says that nephrite was found by the very first Maoris who came to the Island, and it was certainly known in the North Island before the year 1700. We come now to the observational evidence which has been collected during the last forty years. Sir Julius von Haast formed his opinion of the great antiquity of the moa-hunters largely on evidence obtained in the South Island. He ascertained that at the old native cooking-places at the mouth of the Rakaia the ovens of the moa-hunters were only on the higher ground, some 10ft. or 12ft. above the sea, while the Maori ovens which occurred on the lower terraces, about 8ft. and 4ft. above the sea, never contained moa-bones.‡ Trans. N.Z. Inst., vol. iv., p. 96. Again, he showed that the encampments of the moa-hunters east of Christchurch were confined to the inner line of sanddunes, and that only ovens without moa-bones were found on the lower ground north of the Heathcote Estuary, between the sand-dunes and Brighton. Also, at Moa-bone Point Cave, near Sumner, the floor of which was 13 ½ft. above high-water mark, the upper deposits, between 3 ½ft. and 4ft. thick, contained no moa-bones, but only estuarine shells (Venus stutchbryi, Paphia neozelanica, andAmphibola avellana) and Mytilus, while the lower 6in. or 12in. contained moa-bones without estuarine shells. He also showed that these moa-bone beds rested on marine sands, which reached a level of 8 ½ft. to 9ft. above high-water mark, and which contained human remains,§ Trans. N.Z. Inst., vol. vii., p. 74; and McKay, vol. vii., p. 99. as well as bones of seals, and shells of Mactra discors, Paphia spissa, Dosinia australis, and Turritella rosea, all of which are distinctly marine. He also stated that outside the cave, among the old sand-dunes, the shell-fish-eaters lived after the

moa-hunters. All this may be quite true; and yet it does not follow that the moa-hunters were a very ancient race. The explanation is that the land in the neighbourhood of Sumner has been elevated at least 9ft. since the earliest of the moa-hunters lived there. These earliest moa-hunters did not eat shell-fish, because there were none there at that time, the present estuary not being formed until the land rose; and this also accounts for there being no moa-hunter encampments on the flat east of the sand-dunes, this flat being at the time under the sea, and the line of sand-dunes forming the shore. If, also, we suppose that this elevation of the land extended southwards to the mouth of the Rakaia, we get an explanation of the fact that the moa-hunters' ovens only occur on terraces 10ft. or 12ft. above the sea. But Sir Julius von Haast was wrong when, for this reason, he referred these moa-hunters back to Pleistocene times. The elevation of the land 4ft. at Wellington, and 9ft. at the Rimutaka Mountains, on the 23rd January, 1855, is sufficient to show us that we cannot take an elevation of 9ft. as a proof of any great lapse of time. I have elsewhere shown* Trans. N.Z. Inst., vol. viii., p. 103. Fragments of the skull of the sea-elephant (Morunga) were found with the moa-bones. that I consider Dr. Haast to have been mistaken when he thought that he had found at the mouth of the Shag River the same difference between moa-eaters and shell-fish-eaters† Trans. N.Z. Inst., vol. vii., p. 91. as he did at Sumner; for at the Shag River the moa-hunters ate both moas and estuarine shell-fish. Indeed, the recent exploration of Monck's Cave by Mr. H. O. Forbes ‡ Trans. N.Z. Inst., vol. xxiii., p. 373. has proved that the moa still lived near Sumner after the estuary had been formed, for the moa-eggshell, with shellmembrane attached, was associated with estuarine shells and Mytilus, and not with ocean shells as at Moa-bone Point Cave.§ Inside Monck's Cave, as well as just outside, the shells were principally Venus stutchburyi, Mytilus, and Amphibola avellana. Paphia neozelanica also occurred, but not so commonly as the others. The following marine shells are also in the collection from the cave, but I do not know under what conditions they were found: Voluta pacifica, Turbo smaragdus, Haliotis iris, and Pecten laticostatus. The evidence relied upon by Sir J. von Haast to prove the very great antiquity of the moa-hunters no doubt fails; but, on the other hand, are those reasons valid which have been advanced for supposing that the moas lived in the South Island long after they had been exterminated in the North Island? These are, first, the former occurrence of bones lying on the surface of the ground; and, secondly, the discovery of bones with dried skin, ligaments, flesh, and feathers.

With reference to the first: Mr. J. Buchanan, who accompanied the first surveying-party into the interior of Otago, in 1856, says that in the upland district east of the Lammerlaw Range, between 2,000ft.and 4,000ft. above the sea, large legbones of moas were strewn on the surface in great profusion, and they were in very perfect preservation, most of them being quite hard, except where they had been roasted by grass-fires. In the Manuherikia Valley no bones were found on the level terraces, perhaps on account of the late fires; but they were abundant in the scrub on the flats which were liable to be flooded by the rivers.* Trans. N.Z. Inst., vol. v., pp. 416 and 417. The observations of Mr. Vincent Pyke (1861) and of Sir James Hector (1862) I have already quoted, but I may here add that Sir James Hector says that at the south-west extremity of a triangular plain by the side of Lake Wakatipu, in 1862, he counted thirty-seven of such distinct skeleton-heaps.† Trans. N.Z. Inst., vol. iv., p. 118. Mr. Murison (1861) does not mention surface bones. He says, “Searcely a hole could be dug without some of these remains being exposed; and, when the land came to be cultivated, bones and fragments of eggshells in great number were laid bare by the plough.”‡ Trans. N.Z. Inst., vol. iv., p. 121. The Maori cooking-places were also covered by 6in. of silt, and were not discovered until 1865. But all these bones disappeared very rapidly. During the summers of 1873–74 and 1874–75 I rode over the whole of the interior of Otago, making a geological survey, and I never saw moa-bones lying on the surface, except where they had been ploughed or dug up, although they were not uncommon in the river alluvia. The same has apparently occurred in Canterbury. Mr. Boys says that he has seen (date unknown) quantities of moa-bones lying on the surface of the ground on the Waipara Plains.§ Trans. N.Z. Inst., vol. iv., p. 400. While Sir Julius von Haast, who commenced his explorations in Canterbury in 1861, and visited the Waipara in 1866–67, says, “I must confess that I have never observed any [moa-bones] in such positions [i.e., on the ground among the grass on the plains, or between rocks and debris in the mountains], except when it could be easily proved that they had been washed out either by heavy freshes from older deposits in cliffs along river-beds, or by the disappearance of the luxuriant virgin vegetation, consisting of high grass or bushes, the soil having been laid bare, so that its upper portion would speedily be washed away by rain-water.”∥ parallel Trans. N.Z. Inst., vol. iv., p. 70. And a little further on he says that none are found on the surface now

(1871). In 1884 Mr. F. R. Chapman found the remains of nine birds lying on the surface near Lake Tekapo, each with its gizzard-stones, and one with tracheal rings; but he says that these skeletons had been covered with sand, and lately exposed by the wind.* Trans. N.Z. Inst., vol. xvii., p. 175. But bones lying on the surface are not confined to the South Island. According to Dr. von Hochstetter, many years ago numbers of moa-bones were found on the surface near Lake Tarawera, after the forest had been burnt.† “New Zealand,” p. 64. Sir James Hector found bones on the surface in the Raukawa Bush, Hawke's Bay, ‡ Trans. N.Z. Inst., vol. xxi., p. 318. and gizzard-stones have been found with bones in the same district.§ Hamilton, l.c., vol. xxi., p. 319. No doubt many of the surface bones seen by the earlier settlers had been washed out of alluvial beds and brought down to the plains by floods; but this will not account for the more or less complete skeletons, which must have decayed where they were found. Some of these may have been buried in sand for many years and afterwards exposed by the wind, as was the case with those seen by Mr. Chapman. Others may have been covered by dense vegetation, and so protected from the sun, which destroys bones rapidly when they are exposed to its direct rays. But it is, no doubt, difficult to account satisfactorily for all the statements made. One thing however, is clear. If all the bones that were lying on the surface in Otago in 1861 disappeared in fifteen years, either some great change must have taken place in the district during the interval, or else none of the surface bones of 1861 were more than fifteen years old. In the latter case we must suppose that moas were living in large numbers in that district in 1846, three years before the settlement of Otago, although not a single bird was found alive by the first explorers. This is incredible, and we must necessarily fall back on the first suggestion, and account for the disappearance of the bones by the constant burning of the scrub by Europeans, in which case the surface bones do not prove the late existence of the moa in the district where they were found. But it is in the South Island only that bones with dried skin and ligaments have been found. In January, 1864, the specimen of D. potens now in the York Museum was found at Tiger Hill, in the Manuherikia Valley; and, although buried under 14ft. of sand, some portions of the skin and ligaments still remained. ∥ parallel Hector, Proc. Zool. Soc., 1865, p. 751; Allis, Proc. Linn. Soc., 1864, p. 50paeek; Owen, Ext. Birds ofN.Z., p.248, as D. robustus. In 1871 Mr. W. A. Low found a piece of

dried skin, with feathers, in the Dunstan district; * Trans. N.Z. Inst., vol. iv., p. 114. and in the same year the neck of S.crassus now in the Otago University Museum was found in Earnscleugh Cave, near Alexandra.† Trans. N.Z. Inst., vol. iv., p. 115. With the moa-remains were found remains of an extinct duck (Anas finschi), also with dried ligaments; Cnemiornis; and the remains of an extinct lizard, about the size of the tuatara, but with pleurodont teeth (I.c., vol. vii., p. 139)—perhaps one of the lizards mentioned by Mr. Stack (I.c., vol. vii., p. 295). In 1874 the leg of D.potens, also in the Otago Museum, was found by Mr. Allen in the Knobby Ranges.‡ Trans. N.Z. Inst., vol. vii., p. 266; Nature, Feb. 11, 1875; Buller, Birds of New Zealand, 2nd ed., vol. i., p. xxxi., and woodcut from ‘L a Nature.’ In 1878 Mr. Squires found near Queenstown the leg, head, and neck of M. didinus, now in the British Museum; § Trans. Zool. Soc., vol. xi., p. 257. and in 1884 Mr. Brandford discovered the leg of E. elephantopus which is now in the Cambridge University Museum, in a cave in the Remarkable Mountains, near Queenstown.∥ parallel Buller, Birds of N.Z., 2nd ed., vol. i., p. xxxii. Certainly it does not seem probable, at first sight, that these remains can be very old; but, under exceptional circumstances, skin, cartilage, and tendon are known to have been preserved in other places for many hundreds of years.** See Haast, Trans. N.Z. Inst., vol. vii., p. 97, and “Geology of Canterbury,” p. 442. Now, it must be noticed that all the specimens just mentioned have been found in a limited district in central Otago, about sixty miles long and forty miles broad, which lies between Lake Wakatipu and the Lammerlaw Range. So that either the birds survived much longer in this district than in other places,†† The supposed footprints referred to by Hochstetter in “New Zealand,” p. 191, footnote, was a joke perpetrated by Mr. Maling, a surveyor, who had assisted Dr. Haast in digging out the caves near Collingwood. or the remains have been better preserved here than elsewhere. The fact that those bones of the Knobby Range specimen which were exposed to the sun were as much decayed as ordinary moa-bones found on the surface, makes the latter supposition the more probable one; and there are other reasons for coming to the same conclusion. If these remains have not been preserved under special circumstances the birds cannot have been dead more than a few score years at the most: but if they had been alive fifty years before they were found it is certain that the Maoris of Canterbury and Otago would have known that the moas survived longer in central Otago than elsewhere; and yet there is no tradition to that effect. On the other hand, the district in which all these

remains have been found is shut in from the sea on all sides by ranges of mountains, and is, in consequence, by far the driest in New Zealand. Mr. Vincent Pyke, who was Secretary for the Otago Goldfields in 1863, speaking of the remarkable preservative powers of the dry air in this district, says, “On one occasion I was called upon to hold an inquest on the body of a child which was identified as having been the subject of a previous inquest before myself some weeks previously. It had been exhumed from the grave, and appeared slightly mummified, but was otherwise as sound as the day it was buried. On another occasion a boy drowned in the Clutha River on the 1st January was picked up in the following March on a sandy beach 12ft. above the then level of the river, slightly covered with drift sand, but quite fresh and undecomposed, although it had for so long a period been lying exposed to the fierce sun of an extremely hot Dunstan summer.”* “The Moa,” Wellington, 1890, p. 5. Also, in about 1869 the remains of a Maori baby were found in a rock shelter in the same district. It was shown to me in Dunedin in 1872. The skin, hair, ligaments, and some dried-up flesh still remained on the bones. This baby must have died previous to the breaking-out of the diggings in 1861, and probably much before that, for no Maoris are known to have lived in the district or to have visited it since the colony was founded. Skin and ligament, once dried, and protected from the sun, might easily, in this district, be preserved for centuries; so that these remains prove nothing. Conclusion. The case seems to me to stand thus: In the North Island we have, at Wanganui and near Whangarei, undoubted proofs that the ancestors of the present Maoris killed and ate moas. The present generation knew the bones to be those of a bird which they called “moa;” and there are several names of places and of men in which the word “moa” occurs, and these certainly point to a knowledge of the bird. But in the large number of ancient Maori tales and poems which have been collected and published the allusions to the bird are very slight and obscure, and one very ancient poem mentions the moa as having been exterminated before the poem was composed. The so-called traditions of its habits and appearance may be, in large part, later deductions from these words and phrases; and we must conclude that in the North Island the moa was exterminated by the Maoris not very long after their arrival in New Zealand—that is, not less than four or five hundred years ago.

In the South Island, in addition to the proofs that the Maoris killed and ate moas, remains have been found which give the impression that the birds had lived not many years ago. These remaims, indeed, are so fresh that if the birds died fifty years ago they must have been preserved under specially favourable conditions. But certainly the birds have been dead for more than fifty years; while we have reasons for thinking that the district in which these remains have been found is one specially favourable for preserving them. This being so, we cannot say for how many years these remains have been preserved—perhaps for centuries; and, as we have every reason to believe that the ancestors of the Ngaitahu, who have inhabited the South Island for the last two hundred or two hundred and fifty years, never had any personal knowledge of the birds, we must allow that the moas have been extinct for at least that time. On the other hand, it is quite certain that the moa was exterminated by Maoris, and the Maoris are not supposed to have inhabited the South Island for more than five hundred years; so that the time of extinction must fall between these dates. It seems to me improbable that the Ngatimamoe, the last remnant of whom inhabited the West Coast Sounds a few years ago, were moa-hunters. The moa-hunters of the South Island do not appear to have been cannibals, and, as Te Rapu-wai and the Waitaha, who preceded the Ngatimamoe, were, according to tradition, peaceful tribes not given to war, this lends support to the native tradition that it was they who exterminated the moa some three or four hundred years ago—that is, about a hundred years after they had been destroyed on the North Island. Note to Table of Measurements.—In the pelvis the length is that of the pre-acetabular part of the ilium ‘o nly; the breadth is taken at the antitrochanters. The breadth of the sternum is taken across the body just below the costal region. In the skull the length is from just above the foramen magnum to the nasals; B. sq. is the breadth of the squamosals; B. t.f. that at the temporal fossæ; while the height is the vertical from the middle of the basi-temporal.

Table of the Average Measurements of the Species. Metatarsus. Tibia Femur. Pelvis. Sternum. Skull. Length. Girth. Length. Girth. Length. Girth. Length. Breadth. Breadth. Length. B.sq. B. t.f. Height. D. altus 21.5 6.3 D. maximus 20.0 6.5 39.0 8.5 18.5 9.4 D. excelsus 20.0 6.0 37.5 7.0 D. validus 18.5 6.4 35.0 7.0 16.5 8.0 3.8 4.7 2.8 2.2 D. giganteus 18.0 6.0 35.0 7.0 16.0 7.3 D. firmus 16.2 5.2 32.0 6.0 14.8 6.7 8.2 D. robustus 16.0 6.0 31.0 6.5 15.0 7.6 10.0 10.0 8.3 3.7 4.2 2.8 2.1 D. ingens 15.0 4.5 28.5 5.2 13.5 6.3 D. potens 14.5 5.3 28.0 6.0 13.5 7.4 8.5 9.5 7.5 3.5 4.1 2.7 2.1 D. gracilis 13.5 4.5 26.0 5.0 12.4 6.2 3.2 3.7 2.6 1.8 D. torosus 12.5 4.4 24.7 5.2 12.0 6.0 8.0 8.3 7.1 3.2 3.7 2.4 1.8 D. struthioides 11.5 4.0 22.7 4.6 11.0 5.2 7.7 6.7 2.9 3.1 2.2 1.7 P. plenus 10.4 3.8 21.0 4.2 9.5 3.6 7.5 6.0 5.2 P. dromioides 10.0 3.6 19.7 4.0 9.6 3.9 2.7 2.4 1.7 1.3 A. didiformis 6.5 3.0 14.0 3.5 8.0 3.6 5.7 5.4 3.7 2.8 2.6 1.7 1.7 A. antiquus 5.5 12.0 3.0 C. geranoides 5.7 3.0 12.4 2.9 7.3 3.5 5.2 5.2 2.4 1.7 1.8 C. curtus 4.8 2.6 10.5 2.6 6.0 3.0 4.0 3.5 2.7 2.4 1.7 1.4 M. didinus 6.8 3.6 14.9 3.8 9.1 4.4 6.4 7.6 4.2 2.7 2.4 1.8 1.6 S. theides 9.3 5.2 21.2 5.3 12.0 6.3 9.0 11.0 S. crassus 8.5 5.0 19.0 4.8 11.0 5.8 8.1 10.1 6.5 3.8 3.4 2.2 2.0 S. casuarinus 7.6 4.3 17.1 4.0 9.8 5.0 7.0 9.0 5.2 3.7 3.2 2.0 2.0 E. elephantopus 9.5 6.5 22.4 6.2 12.3 7.3 7.0 11.7 8.4 2.9 2.8 2.1 1.9 E. ponderosus 8.5 6.0 19.5 5.5 10.5 6.0 6.0 10.0 7.2 2.7 2.7 2.0 1.8 E. gravis 7.4 5.0 16.7 4.2 9.2 5.0 4.5 7.0 6.5 2.4 2.6 1.8 1.7 E. pygmæus 6.0 4.0 13.5 5.0 7.5 4.5

Explanation of plates XV.-XVII. Skulls of Moas one-half the natural size. Plate XV. Fig. 1 and 1a. Dinornis potens; after Owen, Ext. Birds of N.Z., pl. lxiv. Fig. 2 and 2a. Dinornis (Tylopterx) torosus; after Owen, Ext. Birds of N.Z. of N.Z., pl. Ixxxii. Fig. 3 and 3a. Palapteryx dromioides; after Owen, Ext. Birds of N.Z., pl. xlv. Fig. 4 and 4a. Anomalopteryx didiformis; after Owen, Trans. Zool. Soc., vol. xi., pl. lii. Plate XVI. fig. 5 and 5a. Cela curtus; after Haast, Trans. Zool. Soc., vol. xii., pl. xxxi. Fig. 6 and 6a. Syornis crassus; after Owen, Ext. Birds of N.Z., pl. Ixxvi. Fig. 7 and 7a. Mesopteryx didinus; after Owen, Ext. Birds of N.Z., pl. lxxviii. Fig. 8 and 8a. Euryapteryx ponderosus; after Owen, Ext. Birds of N.Z., pl. lxxvii. Plate XVII. Bones of Anomalopteryx antiquus, from the North Mole Quarry, Timaru: a, right tibia, anterior aspect; b, left tibia, inner aspect; c, right metatarsus, proximal end; d, left metatarsus, distal end. The three other large fragments imbedded in the matrix are portions of the pelvis.