corners; small, low dorsal fins, the 2nd much larger than the 1st, and each preceded by a short dorsal spine of which only the tip is exposed; the upper and lower teeth one-cusped and dissimilar, the uppers lanceolate and sharply pointed, the lowers rectangular with strongly oblique, laterally directed cusps; the dermal denticles loosely spaced, erect; the blades of the denticles of the trunk region very strongly tridentate, deeply hollowed-out and with narrow, flat-topped rims proximally, and a ventral keel.

The family Squalidae includes those small sharks or dogfishes in which there are two dorsal fins preceded by dorsal spines, and no anal fin. Within this family Bigelow and Schroeder (1948, pp. 450-499) recognise nine genera, these being Oxynotus, Centroscyllium, Etmopterus, Deania, Squalus, Cirrhigaleus, Centroscymnus, Scymnodon and Centrophorus. The genera are keyed out and distinguished as follows: Oxynotus is at once distinct and bizarre in that it is very deep-bodied with high dorsal fins and prominent ventrolateral dermal ridges, so that in section the trunk is subtriangular. Moreover each of the dorsal fin spines arises from midway along the dorsal base and extends anterodorsally to emerge from the anterior margin of the fin. Of the remaining eight genera with slender trunks which are not markedly triangular in section, and with the dorsal fin spines lying along the anterior margins of the dorsal fins, Centroscyllium and Etmopterus differ from the rest chiefly in that their upper teeth have several cusps rather than one. Deania includes long snouted, attenuate squalids in which the preoral length is greater than the length from the centre of the mouth to the origin of the pectorals, and is thus in contract to the shorter snouted, less attenuate Squalus, Cirrhigaleus, Centroscymnus, Scymnodon and Centrophorus in which the preoral length is considerably shorter in the same ratio. Squalus and Cirrhigaleus have teeth which are similar in the two jaws, thus leaving three genera, Centroscymnus, Scymnodon and Centrophorus in which the teeth are dissimilar. It is within this latter group that the species described by Thompson as Centrophorus waitei falls.

Bigelow and Schroeder (1948, p. 451) distinguish Centrophorus from Centroscymnus and Scymnodon because it has the inner pectoral corner angular and more or less produced, a feature shown by Centrophorus granulosus (Bloch and Schneider, 1801), the type species of the genus In a further paper, Bigelow, Schroeder and Springer (1953, p. 223) emphasise the value of this character in subdividing “the considerable group of squaloids that agree in having one-cusped blade-like teeth in both jaws and simple dermal denticles”—in other words, the three genera Centrophorus, Centroscymnus and Scymnodon. These authors also indicate firstly the disadvantages of Garman's (1913, pp. 189-233) classification in which the shape of the pectorals is abandoned as “a primary generic character”, and secondly Fowler's (1941, pp. 222-263) distribution of those species with extended inner pectoral corners between two genera, Centrophorus and Entoxychirus, based on “the relative degrees to which the inner corners of the pectorals are produced and the shapes of the dermal denticles” (Bigelow and Schroeder, 1948, p. 451, footnote 8).

The specimen described by Thompson as Centrophorus waitei has the inner pectoral corners rounded and not at all produced. It is thus evident that this species cannot belong to the g Centrophorus as recognised by Bigelow and Schroeder (1948, p. 451) and Bigelow. Schroeder and Springer (1953, pp. 223-227). Its inclusion must then be within either Centroscymnus or Scymnodon

—which have rather similar facies and in which the inner pectoral corners are rounded. The distinctions between adult specimens of these two genera are small, and depend almost entirely on the nature of the dermal denticles. In Centroscymnus the “blades of the dermal denticles on trunk behind 1st dorsal smooth, with rounded margins, ridged or striate denticles confined to more anterior parts of the body” (Bigelow and Schroeder, 1948, p. 451), while in Scymnodon the “blades of dermal denticles with three to several ridges; with marginal teeth on posterior as well as on anterior parts of trunk” (Bigelow and Schroeder, 1948, p. 451). Although it is possible to separate adult specimens of the two genera on these denticle characters, the position has been complicated by the fact that juvenile specimens of Centroscymnus coelolepis have been found to have dentate denticles on the trunk instead of the evenly rounded denticles which occur in the adult (Bigelow and Schroeder, 1954, p. 47). This feature was first noted by Tortonese (1952, pp. 386-387, Fig. 1) who found strongly tridentate denticles on the sides of a juvenile male of 270 mm. total length, which was in all other respects, referable to Centroscymnus coelolepis. Bigelow and Schroeder corroborate Tortonese' findings and show that in a series of specimens ranging from embryos to an adult female of 1035 mm. total length, there is a gradual transition from dentate to smooth-edged denticles, the denticles which develop later being less and less dentate until the adult condition is reached. Accompanying this transition is an increase in the size of the denticles, and in addition, instead of being loosely spaced as they are in the juvenile, the denticles become more and more packed so as to overlap each other. Further evidence cited by Bigelow and Schroeder indicates the probability that a similar transition from tridentate to smooth-edged denticles accompanied by an increase in denticle size and closer packing occurs with growth in specimens of Centroscymnus owstoni Garman, 1906.

The significance of these findings is that juvenile specimens of Centroscymnus coelolepis could be identified as specimens of Scymnodon because the dentate denticles on the sides of the trunk resemble those of Scymnodon and not Centroscymnus, and there are no other features yet recognised to adequately diagnose these two genera. Actually this has happened, and Bigelow and Schroeder (1954, p. 51) find that the juvenile specimens, 330-462 mm. in total length, on which they based their Scymnodon melas (1953, pp. 233-237. Fig. 5), are no other than juveniles of Centroscymnus coelolepis.

The generic status of the juvenile male on which Thompson (1930) based his Centrophorus waitei cannot therefore be determined readily. Not enough is known of the juveniles of Centroscymnus and Scymnodon to lay down any definite criteria, but the nature of the denticles of Thompson's species, together with the information given above on the changes with growth of the denticles of Centroscymnus coelolepis suggest very strongly that on these characters alone Thompson's species should be included in the g. Centroscymnus. The denticles of the head region of C. waitei are ridged as they are in Centroscymnus coelolepis, and differ from those of the trunk, which in C. waitei lack ridges, are smoothly hollowed-out and strongly tridentate And although the denticles of the trunk of C. waitei vary considerably from their evenly rounded, slightly hollowed-out counterparts in adults of Centroscymnus coelolepis, they do show a close resemblance to those described from juvenile specimens of this species by Bigelow,

Schroeder and Springer (1953, pp. 233-237—under Scymnodon melas), and Bigelow and Schroeder (1954, pp. 47-52, Fig. 2).

It is worthwhile at this point to look for further affinities other than denticle characters, between Centroscymnus waitei and other species of Centroscymnus, and also for distinctions between the characters of this genus and those of Scymnodon. Only four other species of Centroscymnus are known, these being C. coelolepis Bocage and Capello, 1864, C. owstoni Garman, 1906, C. fuscus Gilchrist and von Bonde, 1924, and C. cryptacanthus Regan, 1906 The former two are well known, and excellent descriptions and illustrations are available in Garman (1913—both species), Bigelow and Schroeder (1948—C. coelolepis) and other sources. C. fuscus is known only from the description of the type which was not illustrated. According to Smith (1950, p. 58) the type specimen is lost. C. cryptacanthus also is apparently inadequately known and not illustrated. However, the two well-known species have such uniform generic characterisation which is evident also in the descriptions of C. fuscus and C. cryptacanthus that it is reasonable to use the former alone as a basis for comparing the facies of the genus with that of C. waitei.

The three species thus compared, C. coelolepis, C. owstoni and C. waitei, agree in the following: the trunk moderately slender, compressed and uniformly tapered; the head depressed, but obtusely margined, and with large eyes; the snout bluntly pointed and short; the dorsal fins small, with indistinct origins, and each preceded by a dorsal spine of which only the tip protrudes (concealed in C. cryptacanthus); the 1st dorsal smaller in area than the 2nd dorsal, and brush-shaped, while the 2nd is a little more triangular; the caudal well developed and deep, the distal margin truncate or oblique (very oblique in C. waitei, less so in C. owstoni and convexly truncate in C. coelolepis) and a distinct subterminal notch (not prominent in C. waitei); the pectoral fins with broadly rounded corners, and short so that when adpressed to the trunk they fail to reach the level of the 1st dorsal spine; the pelvics just anterior to the second dorsal and similar in shape to it, the spiracles large, above and behind the level of the eyes, the gill-openings small, vertical and arranged in a horizontal series anterior to the pectoral base; the nostrils oblique, with their anteromedial margins expanded as small triangular lobes, the mouth broad, little arched, and bounded by long, deeply incised, upper labial furrows which reach ½ to ⅗ of the distance between the angles and the symphysis of the upper jaw, and short, shallow, lower, furrows; the upper teeth erect, each with a sharply pointed lanceolate cusp arising from a bifid base, and the lower teeth rectangular, and each with a strongly oblique, laterally directed cusp.

The very close agreement which the species of Centroscymnus, including C. waitei, show in the above features, cannot be stressed too strongly at this stage.

The g. Scymnodon includes nine species as recognised by Bigelow, Schroeder and Springer (1953, pp. 230-233), other than C. waitei which they also provisionally regard as a Scymnodon. Most of these species are inadequately known, and only a few are well illustrated—viz., S. foliaceus (Gunther, 1877) and S. squamulosus (Gunther, 1877) in Gunther (1887, Pl. II) and S. jonssonii (Saemundsson, 1922, Pl. V). They do not form as compact an assemblage as do the species of Centroscymnus, so that it is not possible to delineate the facies of the genus as concisely as can be done with Centroscymnus. But the following characters emerge and appear to be valid; the trunk is generally similar to that of Centro-

scymnus, being moderately slender, compressed and uniformly tapered; the head depressed but varying in thickness at the margin, and in this respect generally thinner along the snout than it is in Centroscymnus (as seen in S. squamulosus and S. jonssonii the head appears to have a thin well-defined margin anterior to the eye); the snout bluntly pointed, but narrower and more elongate than in Centroscymnus; the dorsal fins either similar to those of Centroscymnus with a small brush-shaped 1st dorsal and a larger triangular 2nd dorsal in which the base is constricted so that in lateral view the base is distinctly narrower than the distal portion of the web (S. squamulosus, S. jonssonii and S. plunketi), or else with two large triangular dorsals resembling those of the g. Squalus (S. foliaceus); the dorsal spines small and just protruding (S. squamulosus), or with the 1st dorsal spine small and just protruding, the 2nd spine large and strongly projecting (S. foliaceus), or with both spines well developed and strongly projecting (S. macracanthus); the caudal fin similar to that of Centroscymnus, but squarely truncate (even in the embryo of S. plunketi figured by Waite (1914, Pl. III) the caudal is squarely truncate and thus fundamentally dissimilar to that of the juvenile C. waitei), and with a strong subterminal notch, the pectoral and pelvic fins as in Centroscymnus, though Saemundsson (1922. Pl. V) illustrates S. jonssonii with the pelvic base almost opposite the base of the 2nd dorsal, the spiracles large and above and behind the eyes; the nostrils generally oblique, but almost transverse in S. foliaceus, and with their anteromedial margins expanded as short triangular lobes; the mouth little arched and with labial furrows as in Centroscymnus (S. foliaceus, S. squamulosus), or highly arched and with very long upper labial furrows which reach almost to the symphysis (S. jonssonii); the upper teeth erect, each with a short wide, sharp, awl-shaped or triangular cusp (S. foliaceus, S. squamulosus), or with a long, narrow, sharply pointed, lanceolate cusp (S. plunketi); the lower teeth rectangular, each with a broad, triangular cusp, either erect and symmetrical at the centre of the jaw and oblique and laterally directed at the angles (S. ringens, S. crepidator), or oblique and laterally directed along the whole width of the jaw (S. plunketi).

In the absence of complete and detailed information on so many of the species of Scymnodon, the above account of the facies of the genus must necessarily be tentative, and cannot provide an adequate basis for comparison with Centroscymnus. It can be seen, however, that there are characters of at least some of the species of Scymnodon which differ strongly from those of Centroscymnus. Emphasising the principal ones of these, they [ unclear: ] the narrower, more elongate pointed head, with a definite thin margin anterior to the eyes; the strongly developed and projecting dorsal spines, accompanied by larger, triangular fins in some cases; the highly arched jaws with very long upper labial furrows reaching almost to the symphysis; the short, wide, awl-shaped or triangular cusps of the upper teeth; and the lower teeth with erect, symmetrical cusps at the centre of the mouth and oblique, laterally directed cusps towards the angles.

It is difficult to forecast what the significance of the above will be when a detailed survey of all the known species of Scymnodon is completed, but the indications are that in many diagnostic characters there is continuous gradation between Scymnodon and Centroscymnus. In the meantime, reference to the characters selected above seems to promise possibilities for distinguishing species of Centroscymnus from those speces of Scymnodon in which they are most evident.

The specific distinctness of C. waitei from C. coelolepis, C. owstoni and C. cryptacanthus is clearcut. In C. waitei the tips of the pelvics fall short of the posterior end of the 2nd dorsal base by a distance subequal to the horizontal diameter of the eye. In C. coelolepis the tips of the pelvics extend to midway between the posterior insertion and the posterior free tip of the 2nd dorsal, as in Garman's (1913) and Bigelow and Schroeder's (1948) figures, or at least to the posterior insertion in juveniles as in Bigelow, Schroeder and Springer's (1953) figure under Scymnodon melas. In C. owstoni the tips of the pelvics have the same relationship to the 2nd dorsal as in adult specimens of C. coelolepis, that is reaching behind the 2nd dorsal base as in Garman's (1913) figure, while in C. cryptacanthus the tips of the pelvics reach further back to the level of the posterior free tip of the 2nd dorsal. Comparing the levels of the posterior insertion of the pelvics with the tip of the 2nd dorsal spine, C. waitei has the pelvic insertion well anterior to the tip of the spine, while in C. coelolepis and C. owstoni these reference points are at the same level. No data is available for comparison with C. cryptacanthus, but it seems likely that in this species the pelvic insertion would be posterior to the tip of the 2nd dorsal spine.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Using the key characters set out in Bigelow and Schroeder's (1948, p. 494) separation of the species, C. waitei is allied with C. fuscus and distinct from C. coelolepis, C. owstoni and C. cryptacanthus in that the preoral length of C. waitei and C. fuscus is equal to the distance from the eye to the first gill-opening In C. coelolepis the preoral length is less than, and in C. owstoni and C. cryptacanthus the preoral length is greater than the distance from eye to 1st gill-opening. There appear to be no characters to separate C. waitei from C. fuscus other than the lengths of the dorsal bases compared with the interspace between the dorsals. But the lengths of the dorsal bases cannot be determined with precision in C. waitei because the dorsal fins arise smoothly from the dorsal profile without abrupt elevation. The point of origin selected in this study is the first evident elevation of the dorsal profile from the median dorsal longitudinal groove. Also the length of the dorsal bases and the interspaces between them are liable to variation with growth and size. C. waitei is based on a single juvenile of 318 mm total length, which still bears an umbilical scar. C. fuscus is based on an adult of 1,100 mm. total length and is thus not comparable with the specimen of C. waitei. Smith (1950, p. 58) records the dental formula in C. fuscus as 70/35, which is considerably greater than the [ unclear: ] for C. waitei However, C. coelolepis has dental formulae ranging from 40/40 (juvenile male, 339 mm. total length, under Scymnodon melas Bigelow, Schroeder and Springer, 1953, p. 234) to 58/40 and 70/42 (Bigelow and Schroeder, 1948, p. 497), so that the discrepancies between C. fuscus and C. waitei are not adequate in themselves to indicate specific distinction. Moreover, Smith (1950, p. 58) gives no indication of the source of his information on the dental formula of C. fuscus, which is not recorded in the type description, and notes that the type specimen has apparently been lost.

Although there are no grounds for distinguishing C. waitei from C. fuscus, the lack of detailed information on the latter species may hide distinctions, and until such time as further specimens of both species and particularly C. fuscus become available, it is desirable to regard them as separate species.

The identification of C. waitei as a Centroscymnus gives the first record of this genus for New Zealand or Australian waters, for hitherto it has been known

Text-fig. 2.—Centroscymnus waitei. Holotype. Figs. A-E—Dermal denticles from high on the side at the level of the 1st dorsal fin. Figs. A-C—External views. Fig. D—Lateral view. Fig. E—Section across blade of denticle at level ab in Fig. C. Fig F—External view of denticle from upper lip. Figs. G-H—External and lateral views of denticle from ventral surface of snout. Figs. I-J—External and lateral-oblique views of denticle from interorbital region. (N.B. Figs. A-B are at right angles to the surfact of the skin, and thus directly correspond to the lateral view of D. Figs. C, F, G and I are at right angles to the surface of the blade of the denticle and thus do not directly correspond to the appropriate lateral views.) Mdk, median dorsal keel; Vk. ventral keel; Vt, ventral tooth.

only from the North Atlantic, South Africa and Japan. Its occurrence in New Zealand, like that of Triakis attenuata Garrick, 1954, is a further indication of how imperfect our present knowledge is of the components and range of the deep-water shark fauna in this area.

Centroscymnus Bocage and Capello, 1864.

Squalidae with dorsal spines arising at origins of fins and lying along anterior margins of latter, their tips either exposed or concealed, without lateral longitudinal ridges or precaudal pits; snout in front of mouth much shorter than