Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 64, 1935
This text is also available in PDF
(3 MB) Opens in new window
– 87 –

The Skeleton of a Small Moa, Emeus Huttoni Owen.

[Read before the Otago Institute, November 14, 1933; received by the Editor, November 30, 1933; issued separately, September, 1934.]

This account is founded on a practically complete skeleton of a female individual of this small species of Moa which was obtained by Mr David Teviotdale, preparator in the Ethnographic Department of the Otago University Museum, on December, 1929, on the sandy beach at Wakapatu, in Southland. He writes of its discovery as follows:—

“While Mr A. King, of Orepuki, and myself were searching the Wakapatu Beach for Maori curios we found the skeleton of a small species of moa. It was lying in a compact heap with most of the bones in their proper position. I took no bones but those that were together. We found some fragments of egg shell among the bones. Wakapatu Beach is a dreary expanse of windblown sand, which is steadily advancing inland, killing the bush as it advances. The moa skeleton was lying just about where the outskirts of the bush would have been before the sand commenced shifting. I noticed scattered and broken portions of other skeletons lying at intervals in the neighbourhood; but I did not pay much attention to them—the majority were, I think, of a larger species than this one. I found no undoubted evidence of moas having been used for food in the middens here. A few days later, on the sandy beach on the Colac side of the Riverton rocks, I saw portions of moa skeletons also about the original margin of the bush.”

Mr Teviotdale is an experienced observer, having for some years been accustomed to excavating Maori middens and other sites in the attempt to find the chronological order of the artifacts of the Maori.

The bones are those of a full-grown bird, the epiphyses of the leg bones being wholly ankylosed with the shafts; they are well preserved, though rather brittle, the sand having acted, as is usual, as a fine protector from decay. As egg shells were found among the bones, there is every reason to regard it as a female.

The skeleton is not quite complete, as it lacks the fourth cervical vertebra (if the number 21 is common to all species of moa as is usually believed); the pair of small free ribs belonging to the first cervico-thoracic vertebra; most of the phalanges, though the mid-toe of each foot is entire.

The skull is remarkably perfect, though the delicate bones of the palate and of the maxillary arch are missing.

The total height of the skeleton as set up is 1030 centimetres. But the height of any moa skeleton depends on the accuracy with which the taxidermist has articulated the vertebrae, on the curve

– 88 –

of the supporting iron rod, and also on the angle at which the femur is articulated with the pelvis, for sometimes it is set nearly horizontally, at others it is placed at a larger angle with the tibia.

Nomenclature of the Genus and Species.

Since the earlier and several of the later naturalists described the species from the measurements of the leg bones, and as these measurements are used by Oliver for diagnosing the genera in the first place, I commenced with these, and compared them with the various likely skeletons in the Museum collection. I found that the leg bones agreed well with those of a bird which had been labelled “Anomalornis paruvs.” But on looking up Owen's description of this species I saw at once that an error had been made in attributing our skeleton to it.

But when the skull was studied, both that of the new find and of this old skeleton, I was led to identify them as Hutton's “Mesopteryx didina Owen,” or, as it should now be called, Meionornis huttoni Owen. As Oliver does not allow this generic name, I follow him in naming it Emeus huttoni Owen, although the name Meionornis is antecedent to Lydekker's definition of Emeus.

I here wish to pay a tribute to the patience and labour of Mr Oliver, who has summarised the various measurements of the leg bones as given by previous naturalists for the different species of the Dinornithidae. He has converted the older “inches” into centimetres, and has tabulated the series of measurements so as to render it comparatively easy to identify the genera recognised by him, and for each genus he gives similar tables for the species. Although he is not at one with the various zoologists who have made a study of the moas as to the limits and names of the genera, yet he has rendered a useful service in thus bringing together in a convenient form the results of the earlier researches. His book will probably serve for many years as an authoritative guide to the study of the group. As there has been much overlapping of the genera by various authors in the past, owing to the fact that most of the earlier genera were founded on the character of the leg bones, some using the tibio-tarsus, others the tarso-metatarsus, and only a few correlating these with any certainty with the skulls, there has crept into the study a tremendous amount of confusion. It appears to me, therefore, desirable to follow Oliver, if only for the sake of peace, in this troubled realm of systematies. To begin overhauling the nomenclature and trying to find out what each author supposed to be the limits of the genera, as has been attempted by such competent men as Lydekker and Hutton, will be futile until we have complete skeletons of the various genera so that the skulls may be correlated correctly with the leg bones. Parker's excellent account of the skulls takes little or no account of leg bones, and in many cases he does not attempt to record the species, but calls them by some symbol.

– 89 –

It is a waste of time until we have such skeletons to commence the task of unravelling the tangled skein already much involved by Owen as well as by Hutton.

As descriptions of complete individual skeletons are few, it seems worth while to describe the present one in some detail. Owen described the skeleton of Anomalopteryx parvus in 1883, from a nearly complete skeleton; A. oweni was described by Haast in 1886, and Andrews gave a detailed account of Megalapteryx didinus (tenuipes Lydekker) in 1897.

History of the Species.

Emeus Lydekker, 1891 (nec Reichenbach).

  • Meionornis Haast, 1873.

  • Mesopteryx Hutton, 1892.

Emeus Huttoni (Owen).

  • Dinornis huttonii Owen.

  • Meionornis didiformis (part) Haast.

  • Anomalopteryx didiformis (part) Lydekker.

  • Anomalopteryx didina Lydekker.

  • Mesopteryx didina Hutton.

In the early days, when anything like an entire skeleton rarely met with, the species were characterised by the dimensions of the leg bones.

In 1868 Dr Julius Haast described and measured a number of these bones collected at Glenmark, in Canterbury, and among them were several which he allotted to Dinornis didiformis Owen, but he was able to sort these out into three groups according to size—“largest, middle sized, and smallest”—which he numbered 4, 5, and 6 in his list of species collected.

In 1874 Captain Hutton, then curator of the Otago University Museum, gave an account of the leg bones excavated from the Hamilton Swamp in Central Otago, and gives tables of their measurements. On p. 275 he writes:—

“The bones that I have arranged under the name of D. didiformis belong probably to a new species. The tibia is well marked and quite distinct, but the femur and metatarsus that I have associated with it pass almost into D. casuarinus.”

This, by the way, indicates the “guess work” that almost necessarily occurred in these early days.

In 1879 Sir Richard Owen quotes on p. 430 the above words of Hutton, and comments: “Possibly the Dinornis of the South Island with the tibia characteristic of D. didiformis of the North Island may need to be noted as D. huttonii.”

– 90 –

In 1882 Owen gave an illustrated account of a head and neck and a leg and a foot of a bird with the dried skin and flesh wrapping round the bones. To this he gave the name D. didina.

In 1891, Hutton, in his article on “The Moas of New Zealand,” enumerates the species of moa recognised up to that date. On p. 129 he names one of the species Mesopteryx didinus, and gives as synonyms Haast's numbers 5, 6 skeletons above mentioned, Owen's D. huttonii, and, further, Owen's D. didinus. On p. 130 he writes:—

“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.”

No doubt it should if the two birds are the same. I have italicised the words “this” and “it,” for the implication by Hutton is that Owen had had under his observation the same set of bones in both cases. There is, of course, no evidence that Owen ever saw the bones from Hamilton Swamp, while it is quite certain that Hutton had not seen the remains of D. didinus. Nor does Owen in his account of the latter species refer to his former suggestion about D. huttonii, as would have been expected had he regarded the two as synonymous. There seems little justification for Hutton's supposition that the two birds are identical.

Although Owen gives certain measurements of the tibio-tarsus of his D. didinus, it would be impossible to obtain very accurate figures, as the bones were and still are covered almost entirely by dried skin and muscle. And, as a matter of later knowledge, Dr Andrews (1897), in his account of the skeleton of Megalapteryx tenuipes Lydekker, points out that D. didinus Owen is generically, but probably not specifically, identical with it: though Oliver (p. 42) regards them as synonymous. As will be seen later in this article, the characters of the skull of M. tenuipes, i.e., M. didinus Owen, differentiates it from the present skull, which agrees with Hutton's “didinus.”

Nomenclature of the Genus.

The limits placed by different authors on the various genera are very vague, and the number of genera has been multiplied from time to time and later again reduced, so that synonyms are plentiful. It will be wisest, I think, to follow Oliver's arrangement.*

Owen places his “didinus” in the genua Dinornis, but Hutton included his species in his genua Mesopteryx in 1891, as also did Parker in his memoir on the skulls of the moas (1895), but Hutton in 1894 recognised that Haast's genus Meionornis of 1873 should take precedence. Oliver does not recognise either of these generic names, and allots the species “huttoni” to the older genus Emeus.

[Footnote] * A good instance of the task of deciding the question is to be seen in Archey's article on the skull of “Cela geranoides,” where he discusses the difficulty of arriving at a conclusion as to what genus the specific name “geranoides” should be allotted. Oliver, however, places the species, including Archey's specimen, under the title of Emeus exilis, a bird which Hutton attributed to Euryapteryx!

– 91 –

Oliver includes Hutton's Euryapteryx compacta as a synonym of his Meionornis didinus (Hutton) solely on the similarity in leg measurements.

The Description of the Skeleton.

Dimensions of the Leg Bones.

As Oliver uses centimetres in all his measurements, I will follow him here.

Length. Breadth. Middle. Girth at Middle.
Proximal. Distal.
Femur 22.4 8.1 3.2 8.9 11.5
Tibio-tarsus 35.8 9.7 3.5 5.5 9.0
Tarso-metatarsus 16.3 6.2 3.0 7.4 9.0
Percentage of Breadth to Length.
Length. Breadth. Middle.
Proximal. Distal.
Femur 100 36 14.2 39.7
Tibio-tarsus 100 27 9.7 15.3
Tarso-metatarsus 100 38 18 45

The breadth at the proximal end of femur is taken along the axis of the head, as Owen measured it. The dimensions of all the bones were taken between two vertical uprights. Hutton does not mention how he took his measurements; in earlier days, at any rate, he seems to have used a tape, and gives his figures in inches.

The girth is liable to vary with the personal equation, i.e., with the precise point at which it is measured.

As it is on the proportional dimensions of the legbones that several of the genera and species have been founded, and as these bones occur more abundantly than other parts of the skeleton and are readily expressed in numerals, Oliver makes use of these proportions as diagnostic characters and gives a useful “key” to the genera, using the percentage of breadth to length of each of the three bones.

From these figures it will be seen that if we take first the percentages of the breadth at the proximal, middle, and distal extremities in the case of the femur, these numbers—36, 14.2, 39.7—approach most nearly to the figures given by Oliver for the genus Dinornis (p. 35), which are 35, 16, 40.

But the tibio-tarsus of our skeleton has the percentage proportions 27, 9.7, 15.3, which represents a bone of greater stoutness than Dinornis, but agrees well with those of Anomalopteryx, in which the corresponding figures are 27, 9, 14. On the other hand, the tarso-metatarsus in its percentages of 38, 18, 45 indicates either the genus Anomalopteryx (35, 20, 45) or the genus Emeus with 38, 21, 47.

It is here that the value of the skull proves itself, for in the former genus the temporal fossae are usually large and the temporal and lambdoidal ridges are confluent, whereas in Emeus these are

– 92 –

widely separated and the temporal fossae are small (see Parker). Thus, if we had only the legbones before us the bird would have been placed in the former genus, but as we have the skull to take into account, it falls into the genus Emeus as used by Oliver, that is, into Hutton's genus Mesopteryx or Haast's Meionornis.

Leg bones of this species, Emeus (Meionornis) huttoni have been recorded from the following localities in the South Island. It does not appear to have lived in the North Island.

Under the Specific name. Author. Locality. Date. Reference.
A. didiformis Haast Glenmark, Canterbury 1868 Trans., Vol. 1, p. 23
B. didiformis (?) Hutton Hamilton Swamp, Otago 1874 Trans., Vol. 7, p. 278
C. Mesopteryx didina Hutton Enfield, South Canterbury 1892 Trans., Vol. 25, p. 15
D. Euryapteryx compacta Hutton Enfteld 1892 Trans., Vol. 25, p. 11
E. Meionornis didinus Hutton Kapua, Canterbury 1895 Trans., Vol. 28, p. 636

As to D., Oliver regards this species as synonymous; but there is some doubt.

In 1896, p. 559, Hutton brings together the average dimensions of numerous legbones of “M. didina” obtained at Glenmark, Kapua, and Enfield; and a comparison with the dimensions of the present skeleton indicates that it is a small variety of the species.

It occurred to me that this might be a sexual difference, as the presence of egg shells with the skeleton indicates that it is a female. But Hutton (1894, p. 636) writes: “We have only a single point of concentration in each bone showing that there was no difference in size between the sexes.” So perhaps it is a climatic variety, as Southland has a cooler mean annual temperature than Canterbury.

I am loth to form a new species, as we are in ignorance of the range of variation in size for any species of moa. For completion, I add the dimensions of the bones of the old skeleton in this Museum that has masqueraded under the title of “A. parvus.”

Length. Breadth. Middle.
Proximal. Distal.
Femur 23.2 8.6 3.5 9.3
Tibio-tarsus 37. 10.8 3.9 6.
Tarso-metatarsus 18.3 6.3 3.5 7.8*

[Footnote] * The distal extremity is worn and broken.

– 93 –

Oliver (p. 49) gives the following figures for “the type of D. huttonii” from the Hamilton Swamp as having length 38.6, br. 11.4, distal br. 6.3 cm. He refers to Hutton's article in Vol. vii, p. 275, but the table of measurements of “didiformis” are given on p. 278 as the mean length of 7 tibiae is 14.8 inches, prox. width 4.3, and distal width 2.2 inches, which, converted into centimetres, are 37.5; 10.9; and 5.5 respectively. Oliver seems to have made a slip in his figures.

The Foot.

The only complete digit is the middle one of each foot, which measures 126 mm. in length. There is no sign on the tarso-metatarsus of a facet for a fourth digit.

In addition to the complete middle toe of each foot, the basal phalanges of both inner and outer toes of the right foot, and of the inner toe of the left foot, as well as two odd phalanges that do not fit serially, were collected.

There is also one toe with four phalanges which differs in colour from the rest and is somewhat worn as if it has been exposed to the air. It may be the second toe of one foot, but the dimensions are smaller than one would expect; it may have belonged to another individual.

The Skull (Figs 1, 2, 3, 4).

The bird had a small head with a short, bluntly pointed beak, in which the culmen slopes up quite gradually from the tip. In general form it resembles the photograph of Emeus (species not stated) in Oliver's book, p. 47, rather than that of Emeus on pl. lvi of Parker's memoir, for they used the words with somewhat different meanings.

Parker has given so lucid and detailed an account of the moa skull that it is only necessary to refer to a few points which are characteristic of the species.

The roof of the cranium is somewhat elevated between the postorbital processes producing a distinct hillock, which is separated from the convex area between the orbits by a slight depression. Such a hillock, though less pronounced, is seen in Arehey's figure of Cela geranoides and is shown in Parker's silhouettes of “Mesopteryx casuarina” as well as in the species α, β, γ, and especially in the last (which is E. huttoni).

The roof of the orbit makes a continuous circular curve, with well-developed, stout, broad post-orbital process, the anterior margin of which has a slightly forward inclination; its lower end is broad and rounded. The circular curve of the orbital roof is in part due to this forward inclination of the postorbital process. The supra orbital fenestra is oval.

The tympanic cavity is large (much larger than in M. tenuipes, to which this skull bears some resemblance), its upper border makes with its posterior boundary nearly a right angle; and its anterior limit is formed by the pointed zygomatic process.

– 94 –

The temporal fossa is small, does not encroach on to the roof of the cranium; its upper boundary, the temporal ridge, is separated from its fellow by almost the entire breadth of the cranial roof; posteriorly there is a smooth area between the temporal ridge and the anterior lambdoidal ridge which is distinct from the posterior lambdoidal ridge.

The paroccipital processes reach down to the level of the lower surface of the basitemporal; in the figure they appear to extend much further, but this is due to the fact that the photograph is taken rather from above; it is oblique, and not actually a back view.

The rostrum of the basisphenoid is rounded below, the margin of the eustachian groove, which traverses the basi-temporal, ceases before it reaches the mid-line, and here the basitemporal bone is somewhat raised and roughened.

The skull is quite evidently that of a “Mesopteryx,” of which genus Jeffery Parker describes and figures in detail the skull of M. casuarina (pl. lx, fig. 12) and gives silhouettes from different aspects of three species which he terms α, β, and γ; (pl. lxi, lxii). The last species is drawn, he says, from the skull “mounted on the skeleton of M. didina from Hamilton Swamp (Otago University Museum),” p. 378. I was puzzled by the reference, as there is no skeleton so named in our collection; the only possible one was labelled “Anomalornis parvus.” A comparison of this with Owen's account of the species showed at once that this skeleton had been labelled wrongly.

A comparison of the dimensions of the leg bones with those of D. parvus and later with Hutton's “didina” (i.e., D. huttonii Owen), as well as the comparison of the silhouettes of the skull in Parker, proved that the skeleton had thus been wrongly labelled all these years. It is evidently one of the birds referred to by Hutton as “didiformis” from the Hamilton Swamp, obtained in 1874, while he was curator of this Museum. I have been unable to decide whether it is an “individual” skeleton, but I think not. It is fairly complete, though several of the vertebrae are lacking, and the sternum and pelvis are imperfect.

The skull of this species, which as I have noted is Parker's “species γ,” agrees quite closely with the skull from Wakapatu except in its slightly greater length.

Parker speaks of the upper margin of the orbit as being “rightangled in species γ” (i.e., the former “didina”), and so represents it in exaggerated state in the silhouette. An examination of the specimen, however, shows that the supraorbital crest has been cracked in its upper region, and the bone has shrunk apart here; it has been varnished, and is easily mistaken for a natural angular outline; it is really artificial.

None of the skulls of which Parker, on p. 408, tabulates the dimensions has such a short length as 100 mm., but the other measurements are in very close agreement, many being exactly the same as for those given for “didina.”

– 95 –

It is the shorter beak that accounts for the relative shortness of the skull. The nasal processes of the premaxillae had been broken across, but no part of it appears to be missing; and it has been carefully repaired, the edges fit, yet it may be a few millimetres short of its actual length.

The skull of E. casuarinus differs from that of E. huttoni, apart from its greater size, in the following points:—The greater breadth of the postorbital process; the temporal ridge is more marked, being rougher; the supraorbital fenestra is circular; the occipital crest is better developed. The posterior border of the basitemporal is nearly straight, with a vertical drop to the basioccipital, instead of a slope. The paroccipital processes are broader and less pointed. The basisphenoidal rostrum is keeled, not rounded; the basipterygoid processes are relatively longer; the posterior margin of the eustachian groove extends across the basitemporal, then curves backwards and meets its fellow (whereas in E. huttoni the groove ceases before it reaches the median line). At the mesial end of the groove the basitemporal is smooth and flat or only slightly convex (whereas in E. huttoni this area is raised and roughened).

Dimensions of the Skull in Millimetres.

The series of measurements are in accordance with those tabulated by Parker. Those in brackets are additional to his.

Total length 100
Length of basis cranii 30
Length of roof of cranium 67
Width at paroccipital processes 48
Width at squamosal processes 59
Width at temporal fossae 45
Width at postorbital processes 62
Width at preorbital processes 36
Width between the temporal ridges 43
(Width at middle of supra-orbital plates 44)
Height of cranium 41
Width of tympanic cavity 17
Width of temporal fossa 14
Width of orbit 25
Distance between the optical foramina 9
Length of premaxilla 44
Length of body of premaxilla 21
Width of body of premaxilla 20
Length of mandibular ramus 98
Length of mandibular symphysis 13
Width of mandibular symphysis 17
Greatest height of mandible 17
Least height of mandible 8
(Least distance between the supra-temporal ridge and anterior lambdoidal ridge 9)
(Occipital foramen; horizontal diameter 13)
(Occipital foramen; vertical diameter 11)
– 96 –

To complete the list, mention may be made of a much damaged skull which Hutton describes (1896, p. 561) under the title of Megalapteryx tenuipes, which was associated with bones that Hutton attributed to his species. Parker (p. 378, footnote), however, identified it as agreeing very closely with that of Mesopteryx γ (i.e., “didina” of his memoir). It was found in a cave on the Lower Buller River on the West Coast of the South Island.

The few measurements are:—

Length of basis cranii 28 mm.
Width at temporal fossae 38
Height of cranium 33
Distance between the optic foramina 9
Vertical 11
Horizontal diameter of occipital foramen 10

Hutton (1891) also gives a few measurements in inches of the skull of “M. didinus” (p. 130). Taking the higher of the two figures he gives for each of the measurements and converting them into millimetres, we have:—

Breadth at temporal fossae 48.45 mm.
Breadth at post-frontal processes 71.4
Length of lower jaw 108.37
Total length of “head” 127.5

The skull agrees in several respects with that of Cela geranoides described and figured by Archey in 1927. The present skull is 16 mm. shorter, and its height is 2 mm. greater; the length of basis cranii, of roof of cranium, width at paroccipital processes, at temporal fossae and squamosal are almost or quite identical. But the width between the temporal ridges is much less in our skull, less by 11 mm.; the width of the temporal fossa is likewise 11 mm. less in E. huttoni than in Cela geranoides. It has also a shorter premaxilla. But the birds were about the same size.

The Sternum (Fig. 5).

The shape of the sternum is very similar to that of E. casuarinus. The anterior margin is nearly straight, and is, as usual, abruptly bent upwards so that its dorsal edge is here 22 mm. above the level of the under surface of the body of the bone. This margin is produced laterally to form the pre-costal or antero-lateral process on each side, the distal end of which is dilated so as to form a knob, elongated in the antero-posterior direction, which is 20 mm. across its roughened surface.

The costal border, which is 30 mm. in length, presents two facets for the two sternal ribs; they are smooth and separated by a roughened area, and the posterior facet is bounded behind by a roughened prominence. A pneumatic foramen precedes the first facet.

There is really no coracoid pit, there being but a shallow, widely open furrow at the base of the precostal process. The body of the sternum is almost square, but is slightly broader than its length, as

– 97 –

measured from the precostal process to the outer origin of the xiphoid or lateral process; this latter extends outwards and backwards at a slight angle, and its free end is external to a line taken from the tip of the precostal process parallel to the lateral margin of the body. The two xiphoid processes are thus not very divergent. One of them is broken near its tip, but the other is entire and has a truncated end.

The median process, which is really the narrowed continuation of the body, extends further backwards than the xiphoids, and is notched at its extremity, though the sides of the notch are not symmetrical.

Dimensions of the Sternum in Millimetres.

Breadth across the tips of precostal processes (b) 131
Breadth of body at anterior end of costal border (a) 96
Length from anterior margin to tip of median process (d) 176
Length from anterior margin to tip of xiphoid process (e) 160
Length from anterior margin to bottom of notch betwen xiphoid and median process (c) 73
Breadth of median process at level of bottom of the notches (f) 65
Breadth of median process near its tip 23
Distance between outer ends of xiphoids (g) 165
Bieadth of xiphoid near tip 9
Length of costal bolder 30
Length of pre-costal process from pneumatic foramen 35
Bicadth of its distal end 20

The two sternal ribs vary in their length on each side:—

Right side: First rib, 59; second. 80.

Left side: First rib, 58; second, 81.

Figures of the sternum of moas are few in number. Owen (Ext. Birds N.Z.) shows that of E. elephantopus (pl. lxxii), of E. casuarinus (pl. lxxiii, lxxiv), of Dinornis robustus (pl. xcvi), and of D. maximus (xcviii), while Oliver gives illustrations for Anomalopteryx and of Dinornis.

The Pelvis (Figs. 6, 7, 8).

The pelvis is entire except that a small portion of the ventral margin of the pre-acetabular plate of the ilium is broken away; and owing to the brittleness of the bones the ischium and the pubis were broken across on one side during transport, but were repaired without any part being lost.

It presents no peculiarities. The dorsal margin of the is nearly a straight line, formed by a narrow ridge in the preacetabular region where the bones of the right and left sides meet the greatly compressed neural spines of the anterior sacral vertebrae. This ridge extends forwards till it meets the neural spine of the

– 98 –

28th vertebra; the first of the synsacral series. In front of this point the two ilia diverge slightly, and the anterior margin of each bone curves rather abruptly downwards to meet the ventral horizontal margin at an angle which is nearly 90°. (Fig. 6.)

The extreme tip of the ilium is 42 mm. in front of its point of attachment to the 28th vertebra and lies at the level of the 27th.

The outer face of this pre-acetabular region slopes down at a steep angle to its ventral margin, and its surface is somewhat concave Two short ribs, broken short, belonging to the 29th and 30th vertebrae project beyond its margin.

The “iliac ridge” extends back to about the level of a vertical line through the anterior border of the acetabulum; behind this point the ridge divides into two, which diverge above the acetabulum and each is continued along the post-acetabular region, separating the ilium into two areas, an upper horizontal plate (the “pelvicroof” of Owen) and a vertical lamina. (Fig. 7.)

The horizontal plates diverge in an elegant curve and enclose between them the thin plate of bone formed by the expanded ends of the neural spines of the posterior sacral vertebrae (the “sacrum” of Owen) which is 51 mm. across its widest part, and is anchylosed to the horizontal plates on each side. As will be seen from the figure this plate does not reach to the hinder end of the ilia, but tapers to a point. The “pelvic roof” or horizontal plate of ilium, which is 47 mm. wide, is separated from this median plate by a row of holes. The “pelvic roof” plus the “sacrum” is the pelvic “escutcheon.” The outer edge of this portion of the ilium descends abruptly to form a vertical lamina some 40 mm. in height, which is continued to the hinder end of the pelvis.

The ventral border of this plate is gently curved; at first, immediately behind the acetabulum, it is concave, so that some of the sacral vertebrae are visible from the side; further back its margin is convex and completely hides the vertebrae.

The ischium, narrow at its origin, widens posteriorly and terminates at a vertical line through the hindmost end of the ilium. The pubis, though more slender than the ischium, is of greater length, at its origin, below the acetabulum, it is roughened as “pectineal tuberosity,” its hinder end curves upwards rather abruptly and terminates somewhat beyond the end of the other two bones of the pelvis.

It may be worth while to comment of the great difference between the pelvis of the Dinornithidae and the living Ratite birds, in which the “pelvic roof” is either quite narrow, as in Struthio, or is totally absent as in the rest; and the ilia, instead of diverging behind the acetabulum and leaving a wide area occupied by the “sacrum” or plate of bone formed by the expanded neural spines, are here parallel to one another, and the post-acetabular space is either quite narrow, or is absent in Apteryx and Rhea; here the two ilia meet above the sacral vertebrae.

– 99 –

Does this difference indicate a different phylogeny? That it has nothing to do with the walking habit is clear from these differences and from the fact that these two features—a “pelvic roof” and a wide “sacrum”—are present in various flying birds, such as the storks, the parrots, and ducks; birds of widely different habits and relations.

Dimensions of Pelvis.

Pelvis as a whole:—

Total length 350
Greatest breadth across the antitrochanters 165
Breadth across the “escutcheon” 140
Breadth measured ventrally at level of pectineal tubercles 120
Diameter of acetabulum 30


Pre-acetabular length from anterior margin of acetabulum to anterior end of ilium 148
Length from posterior margin of antitrochanter to anterior end of ilium 212
Post acetabular length from posterior margin of acetabulum to posterior end of ilium 170
Greatest height of pre-acetabular region in front of origin of pubis 85
Greatest height of post-acetabular region 40
Breadth of dorsal horizontal plate 47
Distance between the ventral edges at anterior end 80
Distance between ventral edges at posterior end 80


Length from posterior margin of acetabulum 178
Length of free portion, measured from bottom of ilio-ischiac notch 170
Height (width) of free extremity 51
Distance between the distal ends (ventral edges of ischia) 180


Length from anterior margin of pectineal tubercle 235
Length of free portion, measured from bottom of ischio-pubic notch 190
Height (width) of upcurved extremity 190
Distance between the distal ends of the two pubes (ventral edges) 200

The Vertebral Column.

Hutton (1894) states that it is a characteristic feature of the Dinornithidae to possess 21 cervical vertebrae; Andrews (1897) also gives this number for Megalapteryx tenuipes. In the present skeleton only 20 were present, but on articulating the series I found that the 3rd and the 4th do not fit accurately. The facet on the postzygapophysis of the 3rd is elongated, and does not correspond to the prezygapophysis of the next one; the centrum is much too narrow, being 7 mm. across the posterior end, whereas the anterior end of the centrum of the next vertebra measures 10 mm. Hence I must

– 100 –

assume that a vertebra was overlooked in spite of the care taken by Mr Teviotdale in his desire to collect all the bones: even the Atlas and the Axis were gathered up—small bones frequently overlooked by collectors.

The 21 cervical vertebrae are followed by 3 cervico-thoracics (22, 23, 24), with free ribs; three free thoracics (25, 26, 27) and three fixed thoracics fused with the sacrals (28, 29, 30), which have ribs either moveable or anchylosed with their centra.

The synsacrum consists of 18 vertebrae, commencing with the 28th with its floating ribs and ending with the 45th. The 46th is moveably articulated with the preceding, and its transverse processes are not anchylosed to the ilia. I find the same numbers in the old skeleton in the Museum. Hutton (1894) states that the sacrum of this species consists of only 17 vertebrae.

Owen (1875) has given excellent figures of the vertebrae of D. maximus, and Hutton's account is sufficiently detailed, so that it is only necessary here to record any differences in specific characters. He gives details for its ally E. casuarinus. In his general account, Hutton states of the Atlas that it is “broadly oval, the larger diameter being dorsi-ventral,” whereas in this species the horizontal diameter of the neural canal is 10 mm., as against 6 mm. for the vertical diameter.

The Axis, the anterior articulating surface, is transversely oval, its height being about half its breadth. In E. casuarinus the “height is nearly as great as the breadth.”

The odontoid process is small, projecting only 2 mm. from the centrum.

In the 3rd vertebra the neural spine is bifid (noted as being single in E. casuarinus). It rises erectly from the middle of the neural plate, which is almost square, its posterior border being nearly straight.

The 4th vertebra is lacking.

In the 5th and 6th the spine is also bifid, and from the base of each a low ridge passes backwards to be continuous with a low roughened “hyperapophysis” on the outer edge of the post-zygapophysis. To this ridge Hutton gives the name “neural ridge,” which Owen calls the “hyperapophysial ridge.” Hutton has called attention to this “remarkable blending of the neural spine, hyperapophysis, and postzygapophysis” as a characteristic feature of the moas.

In the following vertebrae the same condition holds, so that in the 8th a deep longitudinal trough traverses the neural plate from the bases of the conjoined neural spines to the posterior margin of the neural plate. Here these “neural ridges” are nearly parallel.

Further backwards the distinctness of the spine becomes less, as they diverge from one another and lose in height till they become reduced to mere knobs at the anterior ends of the “neural ridges.” In the 18th, however, the spines again become independent, rising

– 101 –

a little above the ridges, and in the 19th the double neural spine becomes tall, fused basally; in the 20th the two are united along their whole length so as to produce a single spine, which in subsequent vertebrae increases in height.

The centrum bears a hypapophysis in the 3rd to 18th vertebrae. In the 3rd it is stout and single, as in the Axis, but in the 5th it is absent. In the 6th, 7th, 8th, and 9th vertebrae, in place of the hypapophysis, there is a pair of widely separated low tubercles close to the base of the ribs; these are the “parial hypapophyses” of Owen or “catapophyses” of Mivart. These knobs become successively more prominent in the later vertebrae up to the 16th. In the next they move backwards and come to lie behind the rib, and they increase in length so that in the 18th they form definite spines and, moving inwards, become united at the base so as to produce a forked process, the two forks of which in 19–21 become completely fused together to form once more a median hypapophysis; short, almost cylindrical and vertical.

Posteriorly to the 22nd the hypapophysis decreases in height, and its base moves towards the anterior end of the centrum; and once more it is represented in 27th by a mere nodule, which disappears in the 28th.

Another feature is the “inter-zygapophysial canal” of Owen (which Hutton does not describe). In the 3rd there is a small foramen near the middle of the outer margin of the neural plate which is bounded externally by a bony bar, called by Owen the “interzygapophysial bar.” In the 5th the foramen enlarges and is situated further back, and the “bar” is seen to extend from the postzygapophysis to the transverse process.

In the 6th, the “bar” is broader and longer as the transverse process moves outwards from the neural plate so that a broad bridge is formed above the canal, which runs fore and aft between the neural plate and the transverse process. Its anterior opening is between the prezygapophysis and the neural spine; its posterior opening is between the sides of the neural arch and the transverse process. This canal, larger in the earlier vertebrae, decreases in diameter till the 13th, when it is little more than an oval slit. In the next it is a small pore, and in the 16th is reduced to a pit the size of a pin's point.

The structure of the cervico-thoracics and of the thoracic vertebrae present no differences from Hutton's account.

The Ribs.

There are nine pairs of ribs, seven of which are moveable, followed by two pairs of immoveable ribs borne by the two first vertebrae of the synsacrum.

The first rib, borne by the 22nd vertebra, is lacking in our skeleton, and, judging from the capitular facet, it is small; the second rib (23rd vert.) is short, and ends freely. The third is longer, but is also a floating rib. These three constitute the “cervico-thoracic ribs.”

– 102 –

The true thoracic ribs commence with the 25th vertebra—this and the next are attached to the sternum by sternal ribs; the next two (27th and 28th vertebrae) are floating. The ribs of the 29th and 30th vertebrae project beyond the edge of the ilia—the former is broken off about one inch from the pelvic bone; the latter about half an inch beyond it.

The length of the ribs, measured in a straight line from the upper surface of the tuberculum to its distal extremity, is as folows:—

Length of Ribs in Millimetres.
Right side. Left side.
1st Lacking Lacking
2nd 65 60
3rd 130 119
4rth 160 160
5th 177 163
6th 168 162
7th 113* 127

I find in our skeleton of Anomalopteryx didiformis that there are only two pairs of cervico-thoracic ribs (21, 22 vert.); two pairs of complete ribs; three pairs of floating (25, 26, 27); and three belonging to the first three vertebrae of the synsacrum.

Uncinate processes are borne by the 3rd, 4th, and 5th ribs of the 24, 25, and 26 vertebrae; that of the 3rd rib is anchylosed to the rib on both sides, as is that of the 4th of the right side; the rest are not fused. The form of the Uncinate (Fig. 9) may be described as a right-angled triangle, with a rounded apex, a somewhat convexly curved hypotenuse, and a slightly concave vertical side. That of the 2nd complete rib measures 42 mm. in length, with a base of 12 mm. Its shape is much simpler than that of A. didiformis, in which the base is much longer; the upper border starts as a nearly straight line, but then curves upwards to produce a recurved hooked tip. The base is 30 mm., its length is 38 mm., and depth 55 mm., so that, although the skeleton is approximately the same size as that of E. huttoni, the uncinate is much larger. (Fig. 9.)

The Synsacrum (Fig. 8).

The synsacral series consists of 18 vertebrae, namely, the 28th to the 45th inclusive. Hutton gives 17 as the usual condition. The first three of them bear ribs, the first being moveable, the other two fixed to the centrum. That of the 29th exhibits a suture at the capitulum, and its tuberculum articulates with the transverse process just within the edge of the ilium. (In our old skeleton of this species the head of the second rib is moveably articulated with its centrum.) These three vertebrae would be called the “thoraco-sacrals.”

The 31, 32, 33 vertebrae have short, broad, transverse process (? ribs) which are vertically extended as they meet the ilia to which they are of course anchylosed. The process of the 34th is very much

[Footnote] * Is broken.

Picture icon

Fig 1.—Photograph of the skull of Emeus huttom, Natural size. The difference in colour of the bones of the lower jaw and quadrate is due to preservation: they have not been intertered with in any way. As certain outline and ridges do not show up distinctly in the photograph, they have been inked in in this and followin figures.

Picture icon

Fig 2.—Dorsal view of the skull. (Four-fifths natural size.)

Picture icon

Fig 3.—Ventral view of the skull. (Four-fifths natural size)

Picture icon

Fig 4.—Posterion view of the skull (natural size). It is unfortunately not quite truly posed, so that the paroceipital processes appear to reach further down than in reality.

Picture icon

Fig 5.—The strenum, in outline, traced on a photograph (about one-third the natural size). The dotted lines show the various measurements mentioned in the text.

Picture icon

Fig 6.—The pelvis, side view (about one-quarter natural size).

Picture icon

Fig 7.—Dorsal view of pelvis. (X ½.)

Picture icon

Fig 8.—Ventral view of pelvis. The constituent veitebrae of the synsacium are numbered. (X ¼.)

Picture icon

Fig 9.—An uncinate process. A of E. huttoni; B of A. didiformis.

Picture icon

Fig 10.—The complete skeleton

– 103 –

stouter than the preceding or the following and are more ventrally placed, directed backward, dilating distally to abut on the lower wall of the acetabulum. These four are the “lumbo-sacrals.”

The transverse processes of the 35th is very slender, is directed towards the same place, but is almost concealed by the previous process on the right side and entirely so on the left.

The transverse processes of 36, 37 are absent, leaving a gap in the series, the “pelvic fossa.” These two correspond to the “true sacrals.”

The processes of the remaining vertebrae, 38th to 45th, are stout; are expanded as they pass outwards into almost horizontal plates so as to come into contact with one another, or even to fuse, where they abut on the postacetabular region of the ilia; these eight processes are at a more dorsal level than the preceding.

The 46th, though similar in form to those immediately preceding, is, however, not fused with them, and its transverse processes are not anchylosed to the ilia.

The neural spines of most of these “caudosacrals” are expanded so as to form a horizontal plate fused for most of its extent to the hinder part of the ilia, forming the “sacrum” already described.


Andrews, 1897. “On a Complete Skeleton of Megalapteryx tenuipes, Lydekker, in the Tring Museum.” Novitates zoologicae, Vol. 4, p. 188.

Archey, 1927. “On a Moa Skeleton from Amodeo Bay, etc.” Trans. N.Z. Inst., Vol. 58, p. 151.

Haast, 1868. “On the Measurements of Dinornis Bones from Glenmark.” Trans. N.Z. Inst., Vol. 1, p. 23.

—— 1873. Meionronis defined in Presidential Address to Cant. Phil. Inst. Trans. N.Z. Inst., Vol. 6, p. 426.

—— 1886. “On Dinornis oweni, n.sp. of Dinornithidae.” Trans. Zool. Soc., Vol. 12, p. 171.

Hutton, 1874. “On the Dimensions of Dinornis Bones” (from Hamilton Swamp). Trans. N.Z. Inst., Vol. 7, p. 274.

—— 1891. “The Moas of New Zealand.” Trans. N.Z. Inst., Vol. 24, p. 93.

—— 1892. “New Species of Moa.” Trans. N.Z. Inst., Vol. 25, p. 6.

—— 1894. “On the Axial Skeleton of the Dinornithidae.” Trans. N.Z. Inst., Vol. 27, p. 157.

—— 1895. “On a Deposit of Moa Bones at Kapua.” Trans. N.Z. Inst., Vol. 28, p. 627.

—— 1896. “On the Leg Bons of Melonornis from Glenmark.” Trans. N.Z. Inst., Vol. 29, p. 557.

Oliver, 1930. “New Zealand Birds.” Wellington: Fine Arts (N.Z.), Ltd.

Owen, 1875. “On Dinornis (part xxi)—Restoration of the Skeleton of D. maximus.” Trans. Zool Soc., Vol. 10, p. 147. (Extinct Birds of N.Z., p. 391).

—— 1879. “Range of Variation Exemplified in Species of Dinornis.” Extinct Birds of N.Z., p. 427.

—— 1882. “On Dinornis (part xxiii)—A Description of the Skeleton of D. parvus.” Trans. Zool. Soc., Vol. 11, p. 233.

—— 1882. “On Dinornis (part xxiv)—Description of Head and Feet, with their Integuments, of an individual of the species D. didinus. Trans Zool. Soc., Vol. 11, p. 257.

Parker, 1895. “On the Cranial Osteology, Classification, and Phylogeny of the Dinornithidae.” Trans. Zool. Soc., Vol. 13, p. 373.

– 104 –

Explanation of Figures.

  • 1. Photograph of the skull of Emeus huttoni. Somewhat larger than its true size. The difference in colour of the bones of the lower jaw and quadrate is due to preservation; they have not been interfered with in any way. As certain outlines and ridges do not show up distinctly in the photograph, they have been inked in.

  • 2. Dorsal view of the skull.

  • 3. Ventral view of the skull.

  • 4. Posterior view of the skull. It is unfortunately not quite truly posed, so that the paroccipital processes appear to reach further down than in reality.

  • 5. The sternum, in outline, traced on a photograph (about two-thirds the natural size). The dotted lines show the various measurements mentioned in the text.

  • 6. The pelvis, side view (about three-eighths natural size).

  • 7. Dorsal view of pelvis.

  • 8. Ventral view of pelvis. The constituent vertebrae of the synsacrum are numbered.

  • 9. An uncinate process. A. of E. huttoni; B of A. didiformis.

  • 10. The complete skeleton.