Oxynotus bruniensis (Ogilby, 1893). Text-fig. 4, A-I; Text-fig. 5, A-F; Text-fig. 6, A-H.

Although the Prickly Dogfish was listed from New Zealand waters as early as 1890 (Hutton, p. 276, as Centrina salviana Risso) on the basis of a specimen in the Christchurch Museum, and further specimens have been reported as Oxynotus bruniensis (Ogilby) by Thomson (1918, p. 5), Phillipps (1946, p. 18) and Parrott (1958, p. 116), no definitive account of a New Zealand specimen has yet appeared.

The photograph in Thomson is presumably of the Wairau Bar specimen which he records, but this is not stated. This photograph was later used by Waite (1921, fig. 32) in his account of South Australian fishes, again without reference, while in Whitley (1940, fig. 154) it is republished and credited to Waite. Parrott's recent photograph (1958, p. 116) appears to be the only illustration definitely attributable to a New Zealand specimen (other than a newspaper photograph referred to by Phillipps, 1946, p. 19) for no other illustrations have appeared except McCulloch's figure of an Australian specimen in Phillipps (1928, fig. 2).

The Prickly Dogfish is not uncommon in New Zealand waters, and nine trawled specimens (including a pregnant female with well formed embryos) taken between 1953 and 1959 have been available for this study. These were trawled from the east coasts of the North and South Islands, from just north of Cook Strait to as far south as Dunedin.

No other localities are yet known for New Zealand, though the species occurs elsewhere off Southern Australia and Tasmania. Seven of the nine specimens were

taken in 64 fathoms to 140 fathoms; one in 25 fathoms; and for one the depth is not known.

Current recognition of the New Zealand species as Oxynotus bruniensis (Ogilby, 1893) is supported by my material which agrees with the description of the type (Ogilby, 1893, p. 62) and with McCulloch's (1914, Pl. 13) excellent illustration of an Australian specimen. Likewise all my specimens fit the key diagnosis for O. bruniensis in Norman (1932, p. 77) and Bigelow & Schroeder (1957, p. 17) even though they range in size from 525 mm to 722 mm long and show some change with growth of those features used in separating O. bruniensis from the eastern North Atlantic O. centrina (Linnaeus, 1758) and O. paradoxus Frade, 1929. In particular the height of the 1st dorsal spine increases relative to that of the fin, being 45% of the fin height in the 525 mm specimen and 55% in the 722 mm specimen; the 2nd dorsal spine and fin show a similar trend from 57% to 67%. Accompanying these changes is a lengthening of the fin bases relative to the fin heights, the fins therefore becoming less slender and more obtuse. The interspace between the 1st and 2nd dorsal fins also increases slightly relative to the length of the 2nd dorsal base, being 80% of the base in the 525 mm specimen and 95% in the 722 mm specimen. As well as these changes with growth, there is also a considerable amount of variation in proportions and fin shapes between specimens of about equal size; similar variation is reported by McCulloch (1914, p. 81) in his Australian material which included six specimens from 585 mm to 600 mm long.

The key diagnosis of O. bruniensis adapted from Norman (1932) and Bigelow & Schroeder (1957) is as follows:—Spiracle subcircular, its vertical diameter less than half the horizontal diameter of eye; 1st dorsal with the spine sloping a little forwards, and supported by radial cartilages; distance from 1st dorsal spine up to apex of 1st dorsal about equal to length of 1st dorsal spine; distance from 2nd dorsal spine tip to apex of 2nd dorsal about equal to or less than length of 2nd dorsal spine; interspace between 1st and 2nd dorsal fins about equal to length of 2nd dorsal base.

In contrast to O. bruniensis, O. centrina has an ovoid to crescentic spiracle higher than half the length of the eye; the 1st dorsal fin without radial cartilages; a lower 1st dorsal fin, the distance from spine tip to apex only ½ to ⅔ as long as the spine; and a longer dorsal interspace which is not less than 1½ times as long as the 2nd dorsal base. In O. paradoxus the 1st dorsal spine slopes rearwards; the 1st and 2nd dorsal fins are very high, the distance from spine tip to apex at least 1½ times the length of the corresponding spine; and the dorsal interspace about twice the length of 2nd dorsal base.

The dermal denticles of my specimens differ according to the size of the specimen they are from in a manner similar to that reported for O. centrina in Bigelow & Schroeder (1957, p. 14). In my smaller specimens the denticles are mainly tridentate, the median tooth much the largest, and the blades directed steeply away from the skin. (Text-fig. 4, D-E.) They resemble, in these features, the denticles of juvenile Scymnodon plunketi and Centrophorus squamosus. In the denticles from larger specimens, the blades are almost vertical to the skin, and there is a stronger longitudinal keel on the posterior face of the blade; this keel usually has one prominent tooth near its distal end (as do the head denticles of juvenile Scymnodon plunketi), so that most of the blades are four-toothed as a whole (Text-fig. 4, A-C). Occasionally there are two or more teeth on the posterior keel, or the keel may be paired, while other variations include an extra lateral tooth on each side of the blade. The median longitudinal ridge on the anterior face of the blade is also thicker than it is on small denticles, and is multi-ridged, especially near its basal end.

Ogilby (1893, p. 63) and Bigelow and Schroeder (1957, p. 17) record the denticles of Australian specimens of O. bruniensis as having mostly five-toothed blades, but this is not supported by my material where the majority of denticles from the sides of the trunk of mature adults are four-toothed. In a personal communication, Mr. W. C. Schroeder informs me that their record is in error, the specimen they examined actually having four-toothed denticles.

Text-fig. 4. —Oxynotus bruniensis. Fig. A —Dermal denticles from high on side below 1st dorsal of 595 mm male, New Zealand. Fig. B— Anterior view of four–toothed denticle from side of 722 mm female, New Zealand Fig. C—Lateral view of same denticle as Fig. B, showing tooth at distal end of posterior median keel (insets show variation in keel teeth on other denticles from same specimen). Figs. D–E— Anterior and lateral views of three-toothed denticle from side of 525 mm male, New Zealand. Figs. F-I—Outlines of denticle bases from 722 mm female showing four–angled form (Fig. F) and more irregular forms (Figs. G-I), where are additional angles and ridges on anterior margin.

Text-fig. 5.—Oxynotus bruniensis, 560 mm male, New Zealand. Fig. A—Functional upper teeth drawn from complete tooth–bearing membrane removed from jaw and pressed flat; stippled teeth indicate course of row on central sector of arch. Fig. B—Functional lower teeth. Fig. C—First right upper tooth. Fig. D—7th right upper tooth. Fig. E—Diagrammatic plan of upper teeth to show the steeply arched rows, and the longitudinal series. Fig. 7—Scymnodon plunketi, 1,417 mm female (Dom. Mus. No. 2636), plan of upper teeth showing broadly arched rows and radially arranged series.

The upper teeth of O. bruniensis are one-cusped, erect and lanceolate, with about four rows. ^* of teeth functional (Text-fig. 5, A).

Viewed in situ, the teeth occupy a forward–arched quadrangular area, wider than long, at the front of the mouth. However, if the angles of the mouth are incised, a narrow band of smaller, irregular, oblique teeth is displayed on each side of the quadrangular area, trailing rearward and placed on the medial face of the upper jaw towards the angle where they evidently have little function. The arrangement of the teeth on the quadrangular area has recently been suggested by Bigelow & Schroeder (1957, p. 14) as sufficient reason for referring Oxynotus to a separate family Oxynotidae, though they give other reasons also (p. 9). They describe the teeth (p. 13) as “in quincuncial arrangement, the functional rows.

successively longer from front to rear, the first row consisting of three teeth only, the number of teeth greater by one in each successive row, and with about six rows functional.” In my material, I find that to make row counts comparable to those of Bigelow & Schroeder, it is necessary to count the teeth in definite transverse rows, as though the jaw were not arched. Also it is necessary to assume that in moving from one transverse row to the next row behind it, the teeth series making up the second row alternate in position with those of the first row. This method of counting the teeth is very practical in Oxynotus where the teeth of such successive rows are sharply delineated in situ. However, it is not a true indication of the arrangement of the teeth which in fact differs very little from that of other squaloid sharks, having several rows of upper teeth functional. If the entire upper teeth membrane is removed from the jaw and laid flat (Text-fig. 5, A, E), the rows of teeth can be seen to follow an arch around the jaw like that of other squaloids (Text-fig. 5 F shows diagrammatically the row arrangement of Scymnodon plunketi for comparison). The arch in Oxynotus is narrower than is common for most sharks and because of this it is difficult to trace the exact course of the rows at the sides of the mouth. In the central sector of the mouth the rows are well defined, and each series is represented in each complete row.

It can be seen that the morphologically most anterior row of teeth is incomplete at the centre of the mouth, but is represented by the 4th series on each side of the median tooth; similarly the second row lacks a median tooth and sometimes the 1st series on either side of the median tooth also. This incompleteness of the anterior rows is matched in the specimen of Scymnodon plunketi figured here (Text-fig. 5 F) and I have observed it also in Centroscymnus owstonii, C. crepidater and Dalatias licha; it appears to be the usual condition for those squaloid sharks where the upper teeth are well spaced. In contrast such species as Squalus acanthias, S. blainvillii and Deania calcea where the teeth bases overlap or are contiguous, the most anterior row is always complete at the centre of the mouth.

Although the rows of upper teeth of O. bruniensis are not easy to follow at the sides of the mouth, the series are very obvious because of the changing tooth shapes in each series from about the 5th outwards. It is noteworthy that these side series run longitudinally (Text-fig. 5 A, E) as do the more central series. In most squaloid sharks the side series are arranged more or less radially to the arch of the jaw (Text-fig. 5 F, Scymnodon plunketi) but in some, and particularly Dalatias licha the series are more nearly longitudinal as in Oxynotus.

In view of the above, it must be accepted that the arrangement of the upper teeth of Oxynotus is not novel in the Squaloidea and does not merit recognition as a familial character. Much the same can also be said of the other characters which have been suggested as diagnostic of the Oxynotidae. For example, the strong ventrolateral muscle ridges along the trunk have counterparts (admittedly not as strongly developed) in most squaloids. Erect, keeled and toothed dermal denticles are found at least in juveniles of Scymnodon plunketi and Centrophorus squamosus. The chief distinction of the oxynotids appear to lie in the extravagance of these and other features (such as their peculiarly high dorsal fins and long, lanceolate pectorals). Other than this extravagance they have no unique character to separate them from the Squalidae. Consequently recognition of them as a separate family Oxynotidae is largely a matter of systematic convenience.

Seven embryos of O. bruniensis found in my 722 mm female are each about 100 mm long. They were taken in December but although well-formed are attached to large globular yolk–sacs 70 mm wide by 60 mm high, and so are a long way off birth. They have external branchial filaments up to 18 mm long (shown here only in the illustration of the ventral view and only on the left side, Text-fig. 6, G-H). The general appearance is much more like that of a “typical” squaloid than of adult Oxynotus, for their bodies are slender, their dorsal fin margins less