steep-sided, sharp-topped ridge along their length. On most denticles there is also a fourth tooth, smaller than the others, projecting from the anterior end of the median ridge, more or less above the pedicle, with its tip directed vertically away from the blade. A few of the larger denticles have in addition to this fourth tooth, two further tooth like processes, one on each side, arising in a similar fashion from the ridges on the lateral teeth. There may also be an angular process, sometimes pointed, on each lateral margin of the median tooth near its posterior end, though these seem to occur only on large denticles of the largest specimens. In all cases, whether the blades have three simple pointed teeth or have some or all of the additional tooth-like processes mentioned above, they arise from erect, elongate, slender pedicles supported on rhomboidal bases. In the smaller simple denticles, the bases are strengthened by four ridges extending on to the pedicle, one from each angle. In the larger and more complex denticles there is one extra ridge and angle on each side of the anterior margin of the base. Denticles from the side of the trunk below the 2nd dorsal fin are mostly of the simple trifid form, those from the interorbital region include both simple and complex forms, but with the latter predominating; those from the upper lip are thin and leaf-shaped with a single median tooth, and either lacking in sculpture or having the ridges greatly reduced; those from the margin of the lower lip are similarly shaped but thick, almost sessile, with shallow heavy ridges; while those from the caudal axis are ovoid to subcircular in shape, with numerous longitudinal ridges converging at the posterior end where the median tooth is small, and the lateral teeth reduced and irregular in size, shape and number.

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

Head measured to 1st gill-opening 4.9 to 4.7 in the total length. Head strongly depressed, elongate, and shallowly wedge-shaped in profile so that the snout tip is sharply pointed. The least fleshy interorbital distance 2.9 to 3.0 in the head. Snout long, with a well defined thin edge from the leve of the eye to the snout tip. This edge is at the mid-level of the profile between the eye and the nostril, but much nearer to the dorsal profile between the nostril and the snout tip. Contour of snout from above smoothly and rather broadly rounded at the snout tip, but widening rapidly just in front of the nostrils to give a distinct step. Behind the nostrils the contours run parallel as far as the eyes, where they gradually diverge to reach the maximum width of the head at the level of the 1st gill-opening. Eye large, ovoid, twice as long as high, its horizontal diameter 2.0 to 2.2 in the snout. Spiracle large, subtriangular, placed level with the dorsal margin of the eye, and behind it by a distance just less than its own length. Gill openings vertical, concave, and in a horizontal series anterior to the pectoral base. Lengths of the gill-openings subequal, and 2.7 to 3.0 in the horizontal diameter of the eye. Interspace between the 1st and 2nd gill-openings slightly greater than that between the 2nd and 3rd or 3rd and 4th, and about twice that between the 4th and 5th. Nostrils of moderate size, almost transverse, and placed midway between tip of snout and eye. Interspace between the nostrils 1.3 in the prenasal length. Each nostril subdivided into two apertures, one lateral and subcircular, the other medial and ovoid, by the anterior and posterior nasal flaps which are extensions of the anterior and posterior nasal margins. The anterior nasal flap is bilobed, with one pointed process dividing the whole nostril into its medial and lateral portions, and the second pointed process extending over the medial aperture. The posterior nasal flap is thick, fleshy, and internal to the anterior flap. The medial nasal aperture is margined, both anteriorly and posteriorly, by a naked membranous fold. Mouth slightly arched, its width 2.1 to 2.4 in the preoral length, and the latter 1.9 to 1.8 in the distance from snout tip to pectoral origin. The upper lip with a coarsely scalloped margin, the lower lip smooth. Preoral clefts long, reaching 3/5; to ⅔ of the distance from the corners of the mouth to the upper symphysis, and extended posteriorly as furrows for an equal length.

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

Teeth generally similar in both sexes, 14-1-15/14-14 in the 823 mm male, dissimilar in the upper and lower jaws. The upper teeth erect, each with a short awl-shaped, smooth-edged cusp borne on a quadrangular base. The bases of adjacent teeth are in contact but do not overlap each other. Upper teeth at centre of mouth slightly smaller than those on either side of them, while towards the corners of the mouth the teeth also are smaller and in most specimens increasingly oblique. Three or four rows of upper teeth functional at the centre of the mouth, two or three towards the corners. The lower teeth oblique, more so in the females than in the males, each with a smooth-edged, reflexed cusp borne on a subquadrangular base. Each cusp is sharply notched laterally, and overlaps the adjacent cusp so that an almost continuous cutting edge is formed; this is particularly so towards the corners of the mouth where the obliqueness of the cusps increases to the extent where the cusps are almost horizontal. There is no median tooth in the specimens used for this description, though one male examined had a symmetrical median tooth and more or less erect paramedian teeth. The base of each lower tooth slightly overlaps its lateral neighbour. Two rows of lower teeth functional.

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

First dorsal markedly long and low, originating smoothly from the dorsal profile at about the level of the axil of the pectoral. Interspace between the 1st dorsal spine and the tip of the pectoral when the latter is adpressed to the side is slightly more than half the length of the pectoral. Height of 1st dorsal 3.2 to 3.3 in that part of its base measured from the origin of the spine to the posterior insertion of the base, and the latter 1.9 to 1.8 in the head. First dorsal spine short, its exposed portion less than half the height of the fin, scarcely curved, and not reaching to the apex of the fin. The anterior margin of the 1st dorsal, posterior to the spine, broadly convex, the apex not prominent, and the distal margin almost straight. The posterior margin free from the base for a distance more than half the length of the base, and the posterior tip sharply pointed. The 2nd dorsal 1.6 to 1.5 times as high as the 1st, but with a shorter base which (measured from the origin of the spine) is 1.2 to 1.3 in that of the 1st similarly measured. Origin of the 2nd dorsal over the middle or rear part of the pelvic base, with the 2nd dorsal spine originating posterior to the latter. The 2nd dorsal spine very long and curved, its exposed portion 1.3 to 1.4 m the height of fin, its tip level with or just higher than the apex. Anterior margin of the 2nd dorsal fin convex, the apex rounded, and the distal margin straight or slightly concave. The posterior free tip pointed and extending to the level of or slightly behind the epiural origin. Caudal measured from the hypural origin 4.6 to 4.7 in the total length, the epiural lobe moderately developed, the hypural lobe rather weak. Height of the epiural lobe 4.5 to 5.0 m its length and 1.3 to 1.4 in the height of the hypural. The epiural margin almost straight, the terminal lobe truncate or slightly convex and separated from the hypural lobe by an obtuse but distinct subterminal notch. Origin of the hypural a little anterior to or below the level of the posterior insertion of the 2nd dorsal. The lower anterior margin of the hypural straight for most of its length but convex towards the apex which is little more than a right angle and not greatly rounded. The posterior hypural margin very shallowly concave. Pectorals lobate, almost twice as long as broad, the length of their anterior margins 1.8 to 2.0 in the head. The anterior margin concave for the proximal 1/5 of its length but straight or slightly convex distally. The posterior and distal margins convex, the posterior corner smoothly rounded in most specimens but slightly angular in some, and the anterior corner rounded. Pelvics small, subtriangular, the anterior and distal margins straight, the apex smoothly and broadly rounded, and the posterior free tip pointed. The claspers ovoid in section, tapered distally and sharply pointed.

Colour: Uniform light grey on most specimens, but a darker brownish grey on some larger specimens.

Etmopterus baxteri Garrick, 1957. Pl. 26, Figs. A, C.

Since the preparation of my original account of Etmopterus baxteri as a new species based on one specimen from 500 fathoms off Kaikoura, eleven more specimens have been taken. Ten of these are from Cook Strait and were caught during deep-water fishing experiments by members of the Department of Zoology, Victoria University of Wellington. The ten include one male 655 mm long, lined from 780 fathoms in November, 1956; and one male 334 mm long and eight females 261 mm to 382 mm long, taken in a trap set in 600 fathoms in July, 1957. The eleventh specimen, a female of 750 mm, was lined from 480 fathoms off Kaikoura in November, 1958, by Mr. R. Baxter, who had previously caught the type in the same locality.

The additional material, especially the smaller specimens, provides a much better basis for comparison of my species with those from other oceans. This is of particular importance as the type of E. baxteri, a female 742 mm long, was the largest known etmopterid at the time of its description. Examination of the new material now shows that in the extensive size-range represented (261 mm to 750 mm) there is considerable change with growth of those relative propotions formerly thought to be of diagnostic value. The effect of such change is that while larger specimens agree with the criteria for E. baxteri, smaller specimens are akin to E. princeps Collett, 1904, a species known only from the North Atlantic. Similarity between these two species was noted in my previous account, but E. baxteri was thought distinct in having a shorter tail region (length from pelvic origin to tip of caudal equal to distance from snout tip to posterior tip of the 1st dorsal fin in E. baxteri, but reaching to midway between 1st dorsal tip and pelvic origin in E. princeps), and a shorter caudal fin (upper caudal margin equal to distance

from snout tip to 2nd gill-opening in E. baxteri but from snout tip to pectoral origin in E. princeps). It is now evident that with increase in total length the tail region in E. baxteri decreases proportionately relative to the head-trunk length, chiefly because of an accelerated growth rate in the middle region of the trunk. As evidence of the latter, the distance from 1st dorsal spine origin to pelvic origin averages 16.8% (15.5%-18.4%) of the total length in four of my small specimens (261 mm to 290 mm long) but is 20.7% in a 655 mm specimen and 22.4% in the type (742 mm). The head length likewise decreases proportionately with increase of the total length, although to a somewhat lesser extent than the decrease of the caudal fin length. This explains the failure of my diagnosis of a shorter caudal fin relative to the head length in E. baxteri than in E. princeps, for while the length of the upper caudal margin is 18.4% of the total length and less than the length of head to 5th gill-opening (20.7%) in the 742 mm type of E. baxteri, it averages 24.6% in the four small specimens (261 mm to 290 mm) mentioned above where it is slightly greater than the head length which averages 23.8%.

Bigelow & Schroeder (1957, p.63) have suggested further dimensional criteria for separating; E. baxteri from E. princeps, their distinctions depending on the length of the interspace between the 1st and 2nd dorsal fins compared with, firstly, the distance from snout tip to pectoral fin, and secondly, the interspace between the 2nd dorsal and caudal. However, as has alredy been shown, the interspace between the dorsals is a region of accelerated later growth, and in my range of material of E. baxteri, both of Bigelow's and Schroeder's alternative characters are covered, and there is smooth intergradation between them.

I find also that my earlier statement that E. baxteri had “a more conspicuous and differently shaped pelvic flank mark; less oblique nostrils; and upper teeth mostly with 7 or 9 cusps rather than the 5 cusps in princeps” no longer holds true, for all of these features are sufficiently variable in my present material to include both E. baxteri and E. princeps. Thus the pelvic flank mark shows some change with growth, its anterior prolongation which extends over the pelvic base being much more attenuate in small specimens than it is in the large holotype. However, it is similarly attenuate in a 750 mm specimen. The number of cusps on the upper teeth increases with growth, small specimens having mostly 3 or 5 cusps per tooth, and larger specimens 5, or occasionally 7 or 9. Excepting the holotype, the upper teeth of all specimens I have examined have lesser cusps diminishing uniformly in size on each side of the major cusp. The holotype is therefore unusual in having a minute lesser cusp sandwiched between the major cusp and a larger lesser cusu.

The dermal denticles which in mature adults are rather stoutly thornlike are more slender in smaller specimens. Naked areas are also more extensive in smaller specimens, as shown by a 261 mm female where the entire ventral surface of the snout and around the mouth are smooth; all the fins are essentially smooth, including the caudal, which has only a few rows of denticles on the axis and one row along the upper margin; and the naked axillary areas are of greater extent. In a 382 mm female the snout tip, the area between the nostrils and part-way back to the mouth, and around the mouth are still smooth, but the 1st dorsal fin, the caudal fin and the paired fins have at least their proximal halves covered with denticles. In a 655 mm male, only the axillary areas, the tip of snout and the second dorsal fin remain naked.

The density of the denticles also appears to increase with growth, larger specimens having relatively more denticles per unit area than smaller specimens. Accompanying this is a change from an obvious linear arrangement of the denticles along the trunk in small specimens to a random arrangement in larger specimens (Plate 26, figs. A, C). The linear arrangement is largely obscured on the trunk of specimens more than 400 mm total length, but persist on the caudal peduncle and tail even in the largest specimens.

Fig. A —Etmopterus baxteri, juvenile female 261 mm from New Zealand.
Fig. B —E. princeps, juvenile male 262 mm (Mus Comp Zool No. 37457) from western North Atlantic.
Fig C —E baxteri, adult male 655 mm from New Zealand
Fig D —E lucifer, showing pelvic flank mark
Photos M. D. King, Victoria University of Wellington

Despite the similarities which are now apparent between E. baxteri and E. princeps, comparison of two sets of specimens of about equal size does show some differences. The most obvious of these is that the second dorsal spine in E. baxteri is strongly curved and long, reaching almost to the apex of the fin in small specimens, less so in large specimens (Plate 26, Figs. A, C). In E. princeps the second dorsal spine is scarcely curved and reaches only about ⅔ of the distance to the apex in small specimens (Plate 26, fig. B). The dermal denticles of E. baxteri are sparser and more regularly disposed than those of E. princeps (Plate 26, figs. A, B). Differences in proportional dimensions between a 261 mm female E. baxteri and a 262 mm male E. princeps (the latter Harv. Mus. Comp. Zool. No. 37457 and kindly supplied by Mr. W. C. Schroeder of Woods Hole Oceanographic Institution) are definite, the specimen of E. baxteri being considerably narrower in the head, and having smaller and generally more slender paired and median fins. However, these differences could not be sustained entrely in a comparison made by Mr. W. C. Schroeder of a slightly larger male specimen of E. baxteri

with comparable specimens of E. princeps. Mr. Schroeder was able to confirm the differences in the 2nd dorsal spine and in the density of the dermal denticles, but although he found the dorsal fins to be somewhat lower in E. baxteri, other dimensional differences were not apparent. From this it is problematical whether the differences noted in my comparison are due to variation or to specific distinction. This can be solved only by further examination and comparison of both species. As an aid to this I am including here (p. 499) proportional dimensions of six specimens of E. baxteri covering the size-range so far known. In the meantime, the differences in the 2nd dorsal spine and in the density of the denticles seem sufficient evidence for retaining E. baxteri as a separate species.

Etmopterus lucifer Jordan and Snyder, 1902. Plate 26, Fig. D.

Through the courtesy of Mr. K. Fraser, Inspector of Fisheries, Tauranga, and Mr. G. Warman, Kaikoura, I received three additional specimens of an Etmopterus in which the dermal denticles are arranged in definite, uniserial, longitudinal rows, and there is a prominent attenuate pelvic flank mark (Plate 26, Fig. D). These specimens are referable to E. abernethyi Garrick, 1957, a species described from two specimens lined from 100–200 fathorns off Kaikoura, but it is now evident that E. abernethyi is a synonym of the Indo-Pacific E. lucifer Jordan and Snyder, 1902. The additional material shows that the differences which I had previously recognised between E. abernethyi and E. lucifer express changes with growth or variation similar to those noted above for E. baxteri, and do not reflect specific distinction. It should be emphasised that since E. lucifer is one of the smallest species of sharks, all specimens recorded being less than 400 mm long, considerable changes with growth can occur with only small increases in total length. The extent of some of these changes with increments of as little as 22.0 mm in total length is discussed below.

Compared with E. lucifer, E. abernethyi was regarded as long tailed, but in five specimens now available, the length of upper caudal margin ranges from 134% of the interspace between pelvic base and subcaudal origin in the smallest (278 mm long) to 83% and 91% in the largest (341 mm and 396 mm long); in the type of E. lucifer (about 300 mm long) it is about 100% Likewise the interspace between pelvic base and subcaudal origin varies from less than the distance from snout tip to first gill-opening in my 278 mm specimen, to more than the distance to pectoral origin in the 396 mm specimen, while in the type of E. lucifer it is equal to the distance to fourth gill-opening. Thirdly, the snout length which I had thought was longer in E. abernethyi than in E. lucifer ranges from 1.5 times the eye diameter in my smaller specimens to 0.9 times in the larger, and thus overlaps the condition in E. lucifer.

It can be seen that in all three criteria given above the dimensions of the 300 mm type of E. lucifer are about intermediate between those of my smallest (278 mm) and largest (396 mm) specimens. The identification of my material as E. lucifer is confirmed by comparison with five specimens of E. lucifer from Kumanado, Japan (kindly supplied by Dr. T. Abe, East Sea Fisheries Research Laboratory, Tokyo, and Dr. R. Ishiyama, Shimonoseki College of Fisheries). I cannot find any constant differences between my material and these five E. lucifer, either in proportional dimensions, external morphology, dermal denticles, teeth or colour pattern. The colour pattern is somewhat variable, and the mid-dorsal pale band which I had thought characteristic of E. abernethyi occurs also on some of the Japanese specimens. The shape of the black, presumably luminescent areas on the ventral surface of the caudal peduncle varies from that shown in my 1957 (p. 183, Fig. 3C) account to that given for E. lucifer in Bigelow, Schroeder and Springer (1953, p. 272), but most of the specimens approach the former.