Art. XXXII.—On the Skeleton of Notornis mantelli.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 14, 1881, Page 245

Art. XXXII.—On the Skeleton of Notornis mantelli. By T. Jeffery Parker, B.Sc.London; Professor of Biology in the University of Otago. [Read before the Otago Institute, 21st September, 1881.* When this paper was read I had no skeleton of Tribonyx: I have since, however, received two specimens of that bird from Mr. Robbins, of the Hobart Museum, whom I desire to thank for his promptitude in supplying my wants in this respect. The paper has been recast to admit of the necessary comparisons with Tribonyx. January 24, 1882.] Plates XIX—XXI. Introductory. The genus Notornis was founded by Professor Owen in the year 1848, upon portions of the skull sent to him from the North Island by the Hon. Walter Mantell. The skull was fully described in the Transactions of the Zoological Society, and the genus was referred to the family Rallidæ, as a close ally of Porphyrio. Shortly afterwards the same distinguished osteologist received a femur, a tibia, and a tarso-metatarse of the same bird, as well as a sternum, which he at first erroneously referred to Notornis, but afterwards (in 1871) recognized as belonging to Aptornis otidiformis. Professor Owen's description of these bones, published originally in the Proceedings and Transactions of the Zoological Society, are republished at pp. 173 and 199 of his great work, “The Extinct Birds of the New Zealand”: the account of the sternum referred to is on p. 198, and the correction of the position at first assigned to it on p. 340. The memoir on the “Restoration of Notornis” (p. 436) contains nothing new as to the osteology of the genus, and, as far as I am aware, no other descriptions of the skeleton have been published up to the present time. The fossil bones of Notornis mentioned above were all found in the North Island, and the bird was at first supposed to be extinct, but in 1849 the first recorded living specimen was captured on Resolution Island, on the West Coast of Otago, and shortly afterwards a second example on Secretary Island. Both were secured by Mr. Mantell, and are now in the British Museum. Unfortunately in neither case were any of the bones preserved. For thirty years nothing more was seen of Notornis, and it was very generally supposed to have become wholly extinct. But about two years ago, the third known specimen was taken on Captain Hankinson's run on the eastern shore of Lake Te Anau; being run down by dogs in the course of a rabbiting expedition. The captor, Mr. J. Connor, fortunately preserved not only the skin of the bird, but also the dried trunk, and last year forwarded both of them to Dunedin for transmission to England for sale. Through the kindness of Mr. E. J. Spence I was allowed to have the specimens at

the Museum for several weeks, and the notes and sketches of the skeleton then made I am now enabled, by the courtesy of Mr. Connor, to publish. As to the external characters, it happened, most opportunely, that Dr. Buller visited Dunedin while the specimen was in my keeping, and made notes of the skin.* See above, Art. XXXI. It was much to be regretted that the funds of the Museum did not allow of the purchase of these relics, as the desirability of a specimen of this rare example of the New Zealand avifauna being retained in the colony, is obvious. I have, however, through the kindness of two ladies, Miss F. M. Wimperis and Miss Maud McLaren, been fortunate enough to secure for the Museum the next best thing to the actual specimen, namely, two life-sized oil paintings, one outlined from Mr. Gould's figure in Owen's “Extinct Birds,” the other from Dr. Buller's figure, and both coloured from the actual specimen with a fidelity and artistic skill which leave nothing to be desired. As the colouring of the Te Anau specimen differs in some details from that of the British Museum examples, it is a matter of considerable interest to have accurately coloured paintings of it. The latter will not lose their value even if an actual specimen should at some future time be secured. General Description and Measurements. The skeleton consisting, as it does, of the parts saved after skinning (pl. XIX., figs. 1 and 2) is minus the head and the anterior cervical vertebræ, the wing-bones, the bones of the legs with the exception of the femora, and the posterior caudal vertebræ. It is in very good preservation with the exception of the ribs and the femur on the left side, which are shattered, probably by shot, and the right side of the middle xiphoid process of the sternum, which is slightly cut, apparently during skinning. The more important measurements are as follows:— Length of trunk, measured from anterior (dorsal) ends of coracoids to posterior extremity of pelvis 18.5 cm. Length of scapula 8.0 " " coracoid 4.2 " " sternum 6.8 " Width of "(measured just posterior to coracoid grooves) 4.3 " Depth of keel of sternum 0.9 " Length of ilium 10.4 " Width of pelvis at posterior border of acetabula 5.6 " Length of femur 10.3 " For purposes of comparison, however, the absolute dimensions of the parts are of less importance than their proportional dimensions as compared with the corresponding parts in allied genera. I therefore give in

Notornis.

the following table the comparative dimensions of the same bones in Notornis, Ocydromus, Tribonyx, and Porphyrio, taking as a standard the length of the trunk as measured from the centre of a line joining the anterior or dorsal ends of the coracoids to the centre of a line joining the posterior boundaries of the pelvis. This, I think, bears a fairly correct relation to the bulk of the bird, and is independent of the length of either neck or legs. Taking this line as equal to 100, the remaining dimensions maybe taken as percentages of the length of trunk, and thus the proportional size of sternum, pelvis, etc., clearly brought out:— Length of Trunk=100. — Notornis. Ocydromus. Tribonyx. Porphyrio. Length of sternum 36 28 35 40 Width " 24 14 17.5 17 Depth of carina sterni 4.8 4.7 7 13 Length of scapula. 43 35 39 49 " coracoid 22 20 22.5 28 "ilium 56 49 52 43 Width of " 29 21 22.5 21 Length of femur 57 51 49 51 Coraco-scapular angle 97deg; 100deg; 92deg; 86deg; Transverse sternal angle 132deg; 141deg; 122deg; 96deg; Pl. XX. represents outline drawings to scale of the four skeletons, that of Notornis being two-thirds natural size. Vertebral Column. In the vertebral column, the nine posterior cervical vertebræ are left: as in Porphyrio and Ocydromus the total number is fifteen and in Tribonyx fourteen,* That is, in the single specimen of each at my disposal. it is probable that the five or six anterior vertebræ are missing. The last cervical vertebra bears on each side a moveable rib, 3 cm. in length, this being proportionally shorter than the corresponding bone in Ocydromus, Tribonyx, and Porphyrio. The penultimate cervical rib is also moveable, but is short and stout, approaching in form to the anterior (ankylosed) cervical ribs. In this particular Notornis agrees closely with Ocydromus. In Porphyrio and Tribonyx the homologous rib is slender, pointed at its distal end, and fully one-fourth the length of its successor.

Defining, as usual, the first thoracic vertebra as the first in which the ribs articulate with the sternum, there are seen to be seven præsacral thoracic vertebræ, free save for a union of their spines by ossified ligaments. The last, or eighth, thoracic vertebra is ankylosed with the compound sacrum. The same number is found in Ocydromus and Porphyrio, in Tribonyx there are nine thoracic, vertebræ. These vertebræ do not differ in any important respect from those of the allied genera; and even if it were necessary I could not describe them in detail, as it was not possible for me to have the skeleton disarticulated. It is to be noted, however, that the entire thoracic region is proportionally shorter in Notornis than in either of its three allies. This is shown in fig. 3, in which the line xy corresponds pretty nearly with the anterior boundary of the thorax in all four figures, while the line x″y″, which is taken through the anterior end of the pelvis in Notornis, and consequently a little in front of the hinder extremity of the thoracic region, passes in the other three figures well in front of the pelvis. The eight thoracic ribs of each side are flattened bones, divided, as usual, into sternal and vertebral portions; five of the sternal ribs articulate with the sternum, their ventral ends being less crowded together—owing to the greater length of the costal edge of the sternum—than in Ocydromus, but more so than in Porphyrio. Four of the vertebral ribs, the second to the fifth inclusive, have uncinate processes, as in Tribonyx and Porphyrio; in Ocydromus there is an uncinate process also on the first thoracic rib. The position of the uncinate processes in Notornis is similar to that in Ocydromus; they are situated nearer the sternal ends of the ribs than in Porphyrio; in Tribonyx their position is about intermediate. The compound “sacrum” contains one thoracic, five lumbar, apparently four true sacral, and six caudal vertebræ. As far as I can make out these numbers hold good for the other three genera, but in adult specimens it is not easy to decide the exact number of true sacrals. Behind the last ankylosed caudal vertebra come four free caudals; from the analogy of the allied genera these were probably followed by two or three ordinary vertebræ and a pygostyle. Sternum and Shoulder-girdle. The sternum of Notornis, as shown by the above table of comparative measurements, and by pl. XX., is as nearly as possible of the same proportional length as that of Tribonyx, while it is considerably longer than that of Ocydromus, and shorter than that of Porphyrio. Its breadth, proportionally to length of trunk, is considerably greater than in either of the three allied genera. The proportions of the sternum are, however, best seen by reducing it in all four genera to the same absolute length; this is done in figs. 4

To Illustrate Paper On Notornis.

and 4a (pl. XXI.), the former showing the outline of the sternum of Notornis from the ventral face with that of Ocydromus superposed in dotted outline on the left side, and that of Porphyrio in broken outline on the right; while fig. 4a shows the right half of the sternum of Notornis with that of Tribonyx superposed in dotted outline. The sternum of Notornis is broad and flat; its anterior edge is somewhat emarginate, as in Ocydromus, and is devoid of all trace of the manubrium or rostrum (fig. 4, 4a, r) found in Porphyrio, and to a less degree in Tribonyx. The coracoid grooves are even more widely separated than in Ocydromus, instead of having merely the width of the rostrum between them, as in the other genera. The diminution in width of the sternum from its anterior to its posterior end is very gradual; in this respect Notornis most nearly approaches Tribonyx. The external xiphoid processes (e. x. p.) are divergent, not expanded at their distal ends, and are proportionally shorter than in either of the allied genera; the middle xiphoid process is blunt and unossified, the bone terminating in a straight transverse edge, about six mm. from the actual extremity of the process. In this again the resemblance between Notornis and Tribonyx is of the closest kind: the middle xiphoid both in Porphyrio and Ocydromus is completely ossified, terminating in the former by a truncated edge, while it is deeply emarginate in the latter. The keel of the sternum is feebly developed, being hardly deeper, proportionally to length of trunk, than that of Ocydromus. Its anterior edge has nothing of the strong forward convexity found in Porphyrio, but passes almost insensibly into the ventral edge; in this respect the resemblance to Ocydromus would be great, but for the fact that in the latter a strong bifid thickening (fig. 4, k) is formed at the junction of the anterior and ventral borders, whereas the corresponding thickening in Notornis is less marked and shows no tendency to division; the resemblance to Tribonyx is here very marked. Another point connected with the fightlessness of Notornis is the very slight lateral curvature of the sternum; its two sides enclose a dihedral angle (fig. 5 B.) which is nearly as open as that of Ocydromus (A) and considerably greater than in Tribonyx or Porphyrio (C and D). This transverse sternal angle as it may be called, seems to be pretty constantly more open in flightless birds than in the normal members of the same group; its increase, and the correlated diminution of the keel, cause the sternum to approach to the ratite type, as is especially well seen in Didus, Cnemiornis, Stringops, and Aptornis, and to a less extent in Nesonetta, Ocydromus and Notornis. On the whole the sternum of Notornis differs from that of Tribonyx in much the same way as the latter from that of Porphyrio. Tribonyx is, in all important respects a mean between the two extremes furnished by Porphyrio and Notornis. Ocydromus, on the other hand is, in some respects,

intermediate between Tribonyx and Notornis, while in others it is in advance of Notornis, in just the same way as the latter is in advance of Tribonyx, i.e., deviates more from normal carinate characters. In the shoulder-girdle the four genera form a very interesting series: this is shown in fig. 6 (pl. XXI.), in which all four shoulder-girdles are reduced to a common length of trunk, and the coracoids are superposed upon one another, so as to coincide in direction. As regards relative length of coracoid and scapula, Notoris is seen to intermediate between Ocydromus and Porphyrio, Tribonyx having the coracoid a little longer than that of Notornis, the scapula somewhat shorter. The same series is observable in the curvature of the scapula; this is greatest in Porphyrio, next comes Tribonyx, then Notornis, and finally Ocydromus with a scapula nearly straight save at its distal end. A similar gradation is seen in a more important point, namely, in the angle enclosed between the adjacent portions of the coracoid and scapula. As was first pointed out by Professors Huxley and Newton* Huxley, “On the Classification of Birds,” Proc. Zool. Soc., 1867, pp. 418 and 425: Newton, “On the Osteology of the Solitaire,” Phil. Trans., 1869, p. 341, note., one of the most marked features of the Carinatce is the fact that the coraco-scapular angle never approaches 180° as in Ratitee, and is usually less than 90°; the only exceptions mentioned by either author being Didus and Ocydromus, in which the angle is slightly over 90°. This, then, is another morphological character which has a definite relation to the power of flight, the coraco-scapular angle, like the transverse sternal angle, being found, speaking generally, to increase pari passu with diminution of that power. It would, however, be a mistake to suppose that there is anything like a constant relation between flightlessness and increase of the coracoscapular angle. I find, for instance, that it is less in Tetrao than in Vultur, and that of the two skeletons of Stringops in the University Museum, one has the angle less than 90° on both sides, while in the other the angle on the left side is just over a right angle, that on the right being the same as in the other skeleton: so that the angle has undergone little or no increase in a bird in which the carina sterni is practically obsolete, and the furcula rudimentary. But the strangest exception to the rule that the coracoscapular angle in the Carinata is less than 90°, is furnished by that paragon of flying birds, the albatross, in which the angle is fully 100° the same, though to a less degree, is the case in the Nelly (Ossifraga).† That is, measuring by the adjacent portions of the bones only, as in the definition of this angle by Huxley and Newton. Of course if the general direction of the scapula be taken, the angle will be greatly diminished. I may mention, in passing, that the most convenient way to take the coraco-scapular angle, is to trace the outlines of the two bones on a sheet of glass held parallel to the median vertical plane of the body. Thus Diomedea and Ossifraga must be added to the above list of exceptions, as well as Stringops (?), Cnemiornis, Aptornis, Tribonyx, and Notornis. As a very general

To Illustrate Paper On Notornis.

rule, however, a large coraco-scapular angle seems to be correlated with a small carina sterni and large transverse sternal angle, and, taking birds of the same order, there is a tolerably close relation between these structural peculiarities and adaptation to a cursorial life. The table of comparative measurements given above shows* See p.247. that, arranged according to depth of carina sterni, or to size of transverse sternal angle, the four genera of Rallida under consideration must be placed in the following order:–1. Porphyrio, 2. Tribonyx, 3. Notornis, 4. Ocydromus. Fig. 6 shows that the same order is maintained if they are arranged by the coraco-scapular angle, which is least (86°) in Porphyrio, and greatest (100°) in Ocydromus, and it will be seen that the list begins with a good flier (Porphyrio), and ends with a bird of purely cursorial habits (Ocydromus). Similar series may be obtained by turning to other orders; in Anseres, for instance, we have 1. Anser, 2.Nesonetta, 3. Cnemioris; in Columba, 1. Columba, 2. Didus; and in Psittacina, 1. Ara, and 2. Stringops. In all cases loss of the power of flight is associated with the ratite characteristics of increase of transverse sternal and coraco-scapular angles, and decrease of carina. There is still one other point to be observed in connection with the shoulder girdle: if the extremes of our ralline series be compared, i.e., Notornis or Ocydromus with Porphyrio, it will be found that the forward inclination of the coraacoid from its sternal articulation is much less in the flightless forms than in Porphyrio, in other words, that the angle enclosed between the coracoid and a fore-and-aft line drawn through the coraco-scapular articulation parallel to the long axis of the body is greater in Notornis and Ocydromus than in Porphyrio. Such an angle will, of course, vary according to the position of the sternum in respiration, so that its exact size is of no importance and it can only be of use in the comparison of extreme forms. The furcula of Notornis is slender and flattened from before backwards in its median portion; this latter part is, however, very thin, so that the apparent thickness of the bone in a ventral view is deceptive, and as a matter of fact it is nearly as slender as in Ocydromus. As to the form of the furcula, fig. 7 shows that, as in preceding cases, the four genera form an almost perfect gradation, Porphyrio having the thickest and most V-shaped furcula, Ocydromus the slenderest and most U-shaped. Pelvis. In the characters of the pelvis the four genera no longer fall into the same order, the heavy cursorial Notornis having a pelvis of considerably greater dimensions than either of its three allies (pl. XX.): in length, in breadth, and in height the pelvis of Notornis is markedly larger than that of Ocydromus, and very considerably larger than those of Tribonyx and

Porphyrio. In the general form of the ilium and in the relative proportion of præ and post-acetabular regions, Notornis approaches most nearly to Tribonyx, while it deviates, on the whole, in the greatest degree, from Ocydrormis. The excess in size of the pelvis is most marked in its transverse dimensions. This is well shown in figs. 8 and 8A (pl. XXI.), in which the four pelves are drawn of the same absolute length. The ischia and pubes of Notornis are widely divergent; so much so that the pubes can be seen throughout nearly their whole length in a dorsal view. In the other three genera they fall well within the outer boundary of the ilia. This is most pronounced in Ocydromus, in which the obturator notch (ob.) is not seen in a ventral view, being completely hidden by the pubis. In the other three genera, it is well seen internal to the pubis. Thus, arranged according to the characters of the pelvis, the four genera under consideration no longer fall into the same order as when arranged by the shoulder-girdle and sternum. As before, Tribonyx is intermediate between Porphyrio and Notornys, but Ocydromus can no longer be placed in a direct series with the others, since, by the size of its pelvis, it comes between Porphyrio and Tribonyx, while in many of its pelvic characters it goes off on a special line of its own. Summary and Conclusions. To sum up: an examination of the four Rails under consideration shows that flightlessness is accompanied by the following structural peculiarities:— a. The carina sterni is diminished. b. The sternum is widened, and the transverse sternal angle is increased. c. The manubrium, or rostrum sterni, disappears. d. The coracoid grooves recede from the middle line. e. The coracoid becomes more nearly vertical in position. f. The coraco-scapular angle is increased. g. The coracoid and scapula decrease in all dimensions, and the furcula decreases in thickness. h. The pelvis increases in size, and the acetabulum is relatively thrown forward. Professor Owen, from the examination of the fossil bones referred to above, considered Notornis to be intermediate in characters between Porphyrio and Ocydromus. Nothing could have been nearer the truth than this, if we were unacquainted with Tribonyx, to which bird Mr. Gould, from a consideration of external characters, considered Notornis to be most nearly allied. An examination of the skeletons shows that Mr. Gould's sagacity was not at fault. In nearly every respect, Tribonyx is intermediate between Porphyrio and Notornis, approaching more nearly to the latter; the only exceptions

of any importance to this rule are in respect of the length of the scapula and of the sternum, both of which are proportionally longer in Notornis than in Tribonyx, instead of shorter, as they would be if the rule held good throughout. As regards the sternum and shoulder-girdle, Notornis is intermediate between Porphyrio—or rather Tribonyx-and Ocydromus, but in the character of the pelvis it differs far more from the normal ralline type than does Ocydromus. It has always been acknowledged that Notornis is a degenerate rail, its special and aberrant characteristics being due to structural modification connected with the disuse of the wings and the assumption of purely cursorial habits. And I think it is impossible to avoid the conclusion that we have in Porphyrio the nearest living representative of the typical ralline ancestor of Notornis, and in Tribonyx the nearest living representative of an important intermediate stage in the process of degeneration. The degenerate character of Ocydromus is equally evident, but this genus must be looked upon as descended from a strictly ralline, not from a porphyrionine ancestor; in other words, as having come along an entirely different line of descent from Notornis, with which it must have, therefore, only a collateral relationship. I regret that the absence of specimens in this Museum prevents my instituting a comparison with Aptornis, but judging from Professor Owen's figures, it seems to me that this largest of rails is a further development, by degeneration, of an ocydromine type. For instance, the short sternum, narrowing posteriorly, seems a sort of exaggeration of that of Ocydromus, and the pelvis approaches in its proportions to the same genus, having a far less proportional breadth than in Notornis. These conclusions are expressed in the following diagram:—

General Remarks The study of Notornis suggests certain questions of interest as to flightless birds in general, and especially as to that group of preeminently flightles birds, the Ratitæ, and their relations on the one hand to the Carinatæ, and on the other to the reptilian ancestors of the class. The Ratiæ; are usually regarded, from certain undoubted reptilian characteristics, as being nearer the reptilian stock from which birds sprang, than the Carinatæ, and there seems to be a general opinion that large ratite birds of some sort formed the actual connecting link between reptiles of a dinosaurian type and the Carinatæ. But it appears to me that there are serious difficulties in the way of this view. The Ratitæ, in many of the most distinctive avian characteristics, approach no nearer to reptiles than do the Carinatæ. For instance, in the characters of the vertebræ, the femur, the tibio-tarsus, the tarso-metatarsus, the pelvis and sacrum, and, in the case of Dinornis and Apteryx, of the small, backwardly-directed hallux. The chief skeletal characters, except those of the skull, in which the Ratitæ differ from the Carinatæ, are:— a. The absence of a keel to the sternum: b. The great width, and, except in Rhea, the extreme flatness of the sternum, i.e. the openness of the transverse sternal angle. c. The comparatively small size of the scapula and coracoid. d. The coraco-scapular angle is equal to two right-angles. e. The axis of the coracoid is vertical or inclined backwards. f. The furcula is rudimentary or absent. If the Ratitæ are to be looked upon as in any way an ancestral group, these characters must be considered of primary importance, that is, as having a true phylogenetic significance. But in all these points, the Ratitæ merely exaggerate what we find in the flightless members of the Carinate order. There is no more keel to the sternum in Stringops or Cnemiornis than in Struthio, and the transverse sternal angle of Rhea is very considerably less than in the flightless Carinatæ. In these, also, there is a progressive diminution in size of the coracoid and scapula in passing from good fliers to flightless members of the same class, and at the same time a gradual rotation backwards of the dorsal end of the coracoid, and increase of the coraco-scapular angle. In fact, with the exception of the foramen in the coracoid of the ostrich, I know of no character in the shoulder-girdle of Ratitæ which can be pointed out as distinctively reptilian. One important distinction between the shoulder-girdle of reptiles and that of birds, is the position of the bones. In reptiles the coracoid passes from its sternal articulation outwards and slightly upwards, and the scapula (including the supra-scapula), from its

coracoid articulation upwards and somewhat inwards, so that the two pairs of bones form, with the adjacent portions of the sternum and vertebral column, a transverse arch or true shoulder “girdle”: in Carinatæ, on the other hand, the coracoid passes from the sternum forwards, upwards, and outwards, and the scapula, from its coracoid end, backwards and upwards. In the Ratitæ the bones mate no closer an approach to reptiles in this than in other characters, the coracoid is still directed upwards and slightly outwards, the chief alteration in its position being that it has a slight backward inclination, this being, however, only an extreme development of what occurs in Notornis and Ocydromus: the scapula also passes upwards and backwards, and not inwards. Finally, a diminution of the furcula occurs in all birds with functionless wings. In a suggestive paper on the phylogeny of Mammals,* “Nature,” vol. xxiii., p. 227. Professor Huxley has brought out the fact that it gives a wholly erroneous notion of the pedigree of that class to suppose that either the Marsupials or the Monotrermes lie in the direct line of descent of the Monodelphia. He points out that the ancestors of the Monodelphia—the Metatheria—were probably didelphous but not marsupial, and that the Marsupialia are to be looked upon as an offshoot of the Metatheria, which, while retaining the lower characters of brain and urinogenitals, and the large præ-pubes, have undergone great specialization in other directions. In the same way Professor Huxley supposes the Monotremata to be a specially modified offshoot of the Prototheria, the forerunners of the Metatheria. It appears to me that a far juster view of the affinities of the Ratitæ than that alluded to above, is to be had by considering them as the greatly specialized but degenerate (using that word in the sense in which I have applied it to Notornis and other flightless birds) descendants of Carinate birds. Professor Huxley remarks† Proc. Zool. Soc. 1867, p. 419. that “in all probability the existing Ratitæ are but the waifs and strays of what was once a very large and important group.” What I wish to insist upon is that this hypothetical group, like the mammalian Metatheria gave, rise to two races of descendants: one continuing the direct line of descent-the Carinatæ-the other arising by a gradual modification of structure correlated with disuse of the wings-the Ratitæ. Just as the Metatheria gave rise to marsupial descendants which exist now only in the Austro-Columbian and Australian regions, so we may suppose that a widely distributed group of primitive typical birds— Proto-Carinatæ-gave rise to Ratite descendants, now confined to the Austro-Columbian, African, and Australian regions. The fact that such

a group must have been intermediate between Odontornithes* I mean such Odontornithes as Ichthyornis: Hesperornis is another instance of degeneration induced by disuse of the wings. and Carinatæ is quite sufficient to account for all the reptilian affinities of the Ratitæ: the assumption that they had acquired the distinctive characteristics of shoulder-girdle and sternum, pelvis, vertebral column, and fore and hind limbs, and had lost all trace of teeth, will account for the fact that in these points the Ratitæ make no approach whatever to reptiles. The same hypothesis also explains the fact that while the Ratitæ agree in certain common characters, some because they are of ancestral importance, others because they have been acquired by the same law of degeneration, they differ in the most remarkable way in other points. For instance, while Dinornis retains such ancestral structures as the hallux, large external xiphoid processes, and free ischium and pubis, it has undergone the greatest amount of degeneration of the shoulder-girdle and forelimb. Struthio, in the same way, shows the extreme of modification in the foot, Struthio and Rhea in the pelvis, and so on. Probably the great size of the Ratitæ is also connected with their cursorial life, and is, like so many points in the skeleton, merely an exaggeration of what is found in Notornis, Cnemiornis and Didus. The Proto-Carinatæ being, by the hypothesis, good fliers, were presumably not of gigantic size; moreover they probably possessed feathers with connected barbs, so that the special characters of the Ratite plumage should be looked upon as a secondary or degenerate, not as an ancestral, character. Finally, if we look upon the typical Odontornithes and Archæopteryx as approximately linear types in the ancestry of birds, we must assume that the latter arose from ornithoscelidan reptiles of comparatively small size, the gigantic Dinosauria being a special development of the same type. The following diagram expresses the results to which a consideration of the above facts seems to lead:—

Description Of Plates XIX—XXI. Plate XIX. Fig. 1. The Te Anau skeleton of Notornis mantelli, from the ventral aspect (½ nat. size). Fig. 2. " " " " " " from the right side (" "). (Figs. 1 and 2 from photographs by Burton Bros., Dunedin.) Plate XX. Fig. 3. Comparative views of the trunk skeleton of A. Notornis mantelli. B. Tribonyx gouldi. C. Ocydromus australis. D. Porphyrio melanotus. All drawn to the same absolute length. xy, line drawn through anterior boundary of trunk. x‘y’ " " " posterior " " " x“y” " " " anterior end of pelvis of Notornis. x“‘y”’ " " " acetabulum of Notornis. ac, acetabulum. k, keel of sternum. a. tr, anti-trochanter. m.x.p, middle xiphoid process. cl, clavicle. ob, obturator notch. cr, coracoid. pu, pubis. cv-r, last cervical rib. sc, scapula. e.x.p, external xiphoid process. th.r.1, first thoracic rib. il, ilium (præ-acetabular portion). th.r.8 (th.r.9 in B.), last thoracic rib. il', " (post-acetabular portion). uc.p, uncinate process. il. sc, ilio-sciatic foramen. vc, vertebral column. is, ischium. Plate XXI. Fig. 4. Outline of sternum and coracoids of Notornis (even line), with the same parts in Ocydromus superposed on the left side (dotted line), and those of Porphyrio on the right (broken line). For reference letters see under 4a. Fig. 4A. a. Outlines of right half of sternum and coracoids of Notornis (even line) and Tribonyx (dotted line). All drawn to a common length of sternum, and viewed from the ventral aspect. xy, long axis of sternum. cr, coracoid. k, k′, k″, k″′, anterior end of keel of sternum. cp, costal process (Notornis). e.x.p., e.x.p.′, e.x.p″., e.x.p.″′, external xiphoid process. cp′ " " (Ocydromus). m.x.p., m.x.p′., m.x.p″., m.x.p″′., middle xiphoid process. cp" " " (Porphyrio). r, rostrum or manubrium (Porphyrio). cp′″ " " (Tribonyx). r′, " " " (Tribonyx). Fig. 5. Transverse sections of sternum of A. Ocydromus. B. Notornis. C. Tribonyx. D. Porphyrio. All two-thirds natural size. k, keel of sternum.

Fig. 6. Shoulder-girdles of Notornis (thickeven line). Tribonyx (thin " "). Ocydromus (dotted line). Porphyrio (broken "). All drawn to a common length of trunk. cr, cr′, cr″, cr″′, ventral ends of coracoids. sc, sc′, sc″, sc″′, distal ends of scapulæ Fig. 7. Furculæ of A. Porphyrio. B. Tribonyx. C. Notornis. D. Ocydromus. All two-thirds natural size. Fig. 8. Outline of pelvis of Notornis (even line) with that of Ocydromus superposed on the left side (dotted line), and that of Porphyrio on the right (broken line). Fig. 8A. Outline of left half of pelvis of Notornis (even line) and Tribonyx (dotted line). All drawn to a common length of sacrum and viewed from the dorsal aspect. xy, long axis of sacrum. il, ilium (præ-acetabular portion). il′, " (post-acetabular portion). is, is′, is″, is″′, ischium. pu, pu′, pu″, pu″′, pubis. (Figs. 3 to 8 from drawings by the author.)