Art. XX.—On a Specimen of the Great Ribbon Fish (Regalecus argenteus, n. sp.), lately obtained at Moeraki, Otago.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 16, 1883, Page 284

Art. XX.—On a Specimen of the Great Ribbon Fish (Regalecus argenteus, n. sp.), lately obtained at Moeraki, Otago. By T. Jeffery Parker, B.Sc. [Read before the Otago Institute, 12th July, 1883.] Plates XXIII. and XXIV. The genus Regalecus includes a few species of highly specialized deep-sea fishes, which are among the rarest members of the class. Günther, in his “Catalogue of Fishes,”* Vol. iii., p. 307. describes six species, some of which are apparently founded upon single specimens, while some again are rendered decidedly doubtful, owing to the imperfections of the original descriptions and figures, and the absence of further specimens. The total number of examples recorded is very small. Günther states† “Study of Fishes,” p. 522. that only sixteen captures have been made in England between 1759 and 1878. Of these eleven at least were referable to a single species, R. banksii, while one is assigned to R. grillii. Two or three specimens at least have apparently been obtained in Norway (R. glesne), several in the Mediterranean (R. gladius and R. telum), one at the Cape of Good Hope (R. gladius?), one at the Bermudas (R. gladius?), and one at Vizagapatam (R. russellii). In New Zealand a specimen (species doubtful) was found at Nelson in 1860, and described by Mr. W. T. L. Travers. ‡ Günther, “Cat. of Fishes,” iii., p. 307; Hutton and Hector, “Fishes of New Zealand,” p. 35. Another was caught at New Brighton, near Christchurch, in 1876, and was described by Dr. von Haast,¶ Trans. N.Z. Inst., vol. x., p. 246. who made it the type of a new species (R. pacificus). A third was cast ashore on Little Waimangaroa Beach, on the West Coast of the South Island, in 1877; but of this the only description extant§ Quoted by v. Haast, loc. cit. is not exact enough for the

Torpedo Fusca, n. sp. T.J.P. ad.nat.del.

determination of the species. From information received from Mr. G. L. Sise, I have no doubt that a large fish cast ashore near Moeraki about two years ago belonged to the same genus, so that the specimen I am about to describe makes the fifth known to have been found in this country within a period of twenty-three years.* The so-called Regalecus jacksonensis of Ramsay (Proc. Linn. Soc. N.S.W., vol. v., p. 631, pl. xx.) is certainly wrongly named; it is either a large Trachypterus or belongs to a new genus. The species which has been most thoroughly worked out is R. banksii, the subject of an excellent and well-illustrated paper by Hancock and Embleton.† Ann. and Mag. Nat. Hist., 2 ser., iv., 1849, p. 1. R. pacificus is figured and described at some length by von Haast and Powell.‡ Trans. N.Z. Inst., vol. x., 1877, p. 246; and vol. xi., 1878, p. 269. Besides these papers, brief accounts in the works of Günther, Gray, Yarrell, and Cuvier¶ Günther, op. cit.; J. E. Gray, “On the British Specimens of Regalecus,” Proc. Zool. Soc., 1849, p. 80, and Ann. and Mag. N.H., 2 ser., vol. v., 1849, p. 501; Yarrell, “British Fishes,” p. 300; Cuvier, “Règne Animal” (Poissons), p. 148. include all the information at my disposal on the genus. There are, however, numerous references in the “Catalogue of Fishes” to the works of Bloch, Cuvier and Valenciennes, and other foreign ichthyologists. The published figures of Regalecus, like the descriptions, have in many cases been either very incorrectly drawn, or taken from damaged specimens. Nearly all are, however, valuable as showing more or less of such structures as the anterior dorsal and the ventral rays, the precise relations of which in the various species are still extremely doubtful. As specimens may at any time be cast upon our shores, I give in pl. xxiv. outline copies of the head, and, in some cases, of the tail of all the figures I have been able to find: these will, I hope, furnish any local naturalist, who may be fortunate enough to see an example of the great ribbon-fish, with some notion of the chief differences between the species and of the points to be kept in view in making descriptions or drawings. For the same reason the ensuing description will be much fuller than would be necessary in the case of any but an extremely rare and fragile fish. The following is a list of all the figures to which I have found references, those reproduced in the present paper being marked with an asterisk:— 1. Regalecus gladius, Cuvier and Valenciennes, “Hist. Nat. des Poissons.” 2.* " " Cuvier, “Règne Animal” (Poissons), pl. 69. 3. " telum, Cuvier and Valenciennes, op. cit. 4.* " banksii, Hancock and Embleton, loc. cit., pl. 1 and 2 (copied in Yarrell's “British Fishes”).

5. Regalecus banksii, Illustrated London News, June 2, 1849, p. 384. 6.* The so-called Regalecus jacksonensis of Ramsay (Proc. Linn. Soc. N.S.W., vol. v., p. 631, pl. xx.) is certainly wrongly named; it is either a large Trachypterus or belongs to a new genus. " " Gray, loc. cit.: copies of two original drawings in Sir J. Banks's copy of Pennant's Zoology. 7.* The so-called Regalecus jacksonensis of Ramsay (Proc. Linn. Soc. N.S.W., vol. v., p. 631, pl. xx.) is certainly wrongly named; it is either a large Trachypterus or belongs to a new genus. " 8. " " (?) (Gymnetrus hawkinsii), Bloch, “Naturgeschichte der auslœndischen Fische,” xii., pl. 425 (copied by Yarrell, op. cit., p. 302). 9. " " Lütken, “Videnskabelige Meddelelser fra den naturhistoriske Forening” for 1881, p. 190. 10. " glesne, Ascanius, “Icones Rerum Naturalium,” pl. 11. 11. * The so-called Regalecus jacksonensis of Ramsay (Proc. Linn. Soc. N.S.W., vol. v., p. 631, pl. xx.) is certainly wrongly named; it is either a large Trachypterus or belongs to a new genus. " " Schneider, “M. E. Blochii Systema Ichthyologiæ,” pl. 88 (copied by Yarrell, op. cit., p. 301). 12. " " “Encyclopédie Méthodique,” fig. 358. 13. " grillii, Lindroth, Vet. Akad. handl., pl. 8. 14. " russellii, Shaw, “General Zoology,” iv., pl. 28. 15.* The so-called Regalecus jacksonensis of Ramsay (Proc. Linn. Soc. N.S.W., vol. v., p. 631, pl. xx.) is certainly wrongly named; it is either a large Trachypterus or belongs to a new genus. " pacificus, Haast, loc. cit., pl. 7. The specimen from which the following description is taken was cast ashore at Moeraki, about 40 miles north of Dunedin, on the 14th of June. It was purchased by a fisherman, who, previously to bringing it to Dunedin for exhibition, cut it into four pieces for convenience of transit! After it had been exhibited for two or three days I was able to secure it for the Museum. As the specimen was thus wantonly injured, and was moreover by no means fresh when it came into my possession, it was useless to attempt to stuff it, and I decided instead to have the skeleton prepared: of this I hope, as very little seems to be known of it, to publish a detailed description. In the present communication I propose to give a general description of the fish with especial reference to one or two points left more or less uncertain by other observers. All the species of Regalecus are distinguished by their great length in proportion to their height and thickness, most of them being from 8 to 18 feet long, 6–15 inches high, and not more than 2 or 3 inches thick. In the species recorded up to the present time the proportion between height and length varies from 1: 24 (R. telum) to 1: 11 (R. pacificus): in the Moeraki specimen this proportion is as 1: 10, the total length being 12½ feet, and the greatest height 15.25 inches, so that the fish is higher in proportion to its length than any former specimen. The length of the head varies from 1/19th of the total length (R. gladius and pacificus) to 1/16th (R. banksii); in my specimen the head is 1/17th the length of the whole body, being 9 inches long with the jaws retracted. In correspondence with the great length of the fish, the number of rays in the continuous dorsal fin is very considerable: it varies from 134 (8/126) in R. glesne to 406 in R. grillii. In the present specimen the number is

15/190; that is, there are 205 rays in all, the first 15 being lengthened to form a crest, and, in accordance with the usual method of counting, being considered as the first dorsal, the remaining 190 forming the second. In the various descriptions and figures to which reference has been made there is great discrepancy with regard to the number and character of the rays forming this crest or anterior dorsal fin. Hancock and Embleton say of R. banksii (see pl. xxiv., fig. 3):—“The anterior part of the fin, more prominent than the rest, is composed of twelve rays, which were stated by the captors to have been 12 or 14 inches in length when the fish was taken, and to be each furnished with a membranous expansion on its posterior edge, increasing in width upwards something like a peacock's feather. “The first ray is a pretty strong spine arising just within the frontal curve, the three next are very slender, and much closer together than the rest, and when we first saw the fish, united for 4 or 5 inches (their length at that time) by a membrane; the next is equally slender with the preceding, but rather farther apart; the three or four after this are nearly as strong as the first, the rest diminish in strength and length, and become uniform with the rays of the dorsal fin. “It is difficult for us to say whether the twelve front rays constituted a detached crest or formed merely the anterior continuation of the dorsal fin, though after careful and repeated examinations we found shreds of membrane in each interval between them, and their bases also were connected with a continuous membrane. In the interval between the twelfth and thirteenth rays the remains of a membrane were found connecting the bases of these rays, and their shafts were ragged and woolly-looking, as if a membrane had been torn off from them. We are, therefore, inclined to conclude that the crest was really a continuation of the dorsal fin and not a separate structure, though it is probable enough that the ends of its rays may have been for some distance free and even furnished with a membrane on their posterior margin widening at the top, giving them the appearance of peacock's feathers, as asserted by the fishermen. This probability is heightened by the fact of the head of the Gymnetrus [Regalecus] from the Cornish coast being provided with two long rays having broad membranous expansions at their ends, which would justify a casual observer in comparing them in form to the above feathers. It is not unlikely besides that the second, third, fourth, and fifth rays, on account of their resemblance in delicacy to the ordinary fin rays, may have terminated differently to the rest. The rays having been broken we cannot say of ourselves whether they are uniform in size [i.e. length] or not; but by what we have learnt by questioning those who saw the fish, we conclude that the middle rays

were the longest, those in front and behind them gradually decreasing in length. The rays of the crest are more closely set than those of the rest of the dorsal fin, which stand about half an inch apart.” Lütken,* Ann. and Mag. N.H., Ser. 5, vol. xi., p. 181. describing a drawing of a specimen from the Faroe Islands, obtained in 1852, and probably referable to the same species, states that immediately in front of the true dorsal fin “there were two high and pointed nuchal fins, the total number of rays in which cannot be stated exactly,” but is probably about 11. None of these rays exhibit terminal dilatations. Von Haast says of R. pacificus (pl. xxiv., fig. 1) “the first nine spines form a crest. These spines enlarge at their termination to a lobe, as shown by the two only perfect ones when the fish was obtained; they cover a space of 2.5 inches. The first of these spines is broken off at 3 inches from the base; it is the stoutest of the whole series. No. 2 is considerably thinner, and 7 inches long. It is one of the complete ones. The three next spines (3, 4, and 5) were all broken off at 4 to 6 inches, and were nearly as thick as the first. From here they get thinner, the thickness of the seventh having only the thickness of the second. This spine, which is entire, is 7.75 inches long, and has, like the second, a lobe at its termination. The eighth is still thinner, and broken off 1 inch from its base, and there is only a fragment of the ninth, which is not thicker than one of the rays of the dorsal fin proper. All of these spines, which have minute hooks directed upwards on their anterior and posterior edges, are united with each other by a small membrane about .45 inch high. They had, like the two ventral rays, a red colour, very bright in their upper portion when the fish was first obtained, which, however, gradually faded to a dull light pink.” Nothing is said as to the form of the lobes terminating the two perfect spines, but the latter appear from the figure to have been merely bluntly clavate. It is further stated that the proper dorsal fin begins half an inch behind the last of these rays. This statement would seem to imply that the two dorsals were separate in Haast's specimen. In the Nelson specimen the crest is thus described by Travers:—“from the back of the head rose several rigid circular spines, about eighteen inches long, three-quarters of an inch in diameter at the base, tapering to a point, curving slightly backwards, hollow and bristling along their whole surface with small spines directed upwards. These long spines appear to have been very brittle, as they broke off short when the fish struck the rock. The person who saw the fish run ashore described these spines [sic] as presenting the appearance of three small masts to a boat, through the whole length of the fish, disposed in pairs as follows,—one pair just below the back and

Recalecus Arcenteus. n. sp. T.J.P. ad. nat. del.

the other pairs immediately above.” I have italicized the concluding portion of this quotation, as I am totally unable to make it tally with the first part, or indeed to understand its meaning. In R. gladius, Günther states that “the anterior twelve dorsal rays are produced, the first five forming a separate division above the eye, the seven following terminate in cutaneous lobes.” This description is probably taken from Cuvier and Valenciennes and tallies exactly with the figure (see pl. xxiv., fig. 6) in the “Règne Animal,” in which the five anterior rays are united by membrane for fully three-fourths of their length and taper away distally to a point; the seven posterior are only united for a short distance from the base, terminate in lanceolate lobes and show a progressive diminution in length from before backwards, the sixth ray being more than four times the height of the head, the twelfth little more than half that height. The terminal lobes on these seven posterior rays are definitely lanceolate, instead of being formed by a gradual widening of the membranous investment of the spine as in Haast's and Hancock's figures. The Bermudas specimen is similarly described* J. M. Jones, Proc. Zool. Soc., 1860, p. 185. as having “a series of ten or eleven erect quill-like flexile filaments, from 2 to 3 feet in extent, gradually tapering from base to apex, and possessing, in the case of the three longest, lanceolate points.” Of two specimens caught off the Northumberland Coast and described to Hancock and Embleton,† Loc. cit., p. 15. it is said that “there were four processes about 18 inches long from the head, of a red colour, like the feelers of boiled lobsters; they tapered gradually towards their ends, which were enlarged to the form and size of a large button.” These are all the detailed descriptions of this curious crest which I have been able to find. Of the two drawings in the Banksian Copy of Pennant's Zoology, one (see pl. xxiv., fig. 4) shows a crest of eleven rays, all tapering distally and not united by membrane; the length of the first is more than twice the height of the head and it is curved forwards and downwards; the second is barely longer than the height of the head; the rest diminish progressively. The second of these drawings (pl. xxiv., fig. 5) is evidently taken from a specimen in which all the rays except the seventh were so broken as to give no indication that they were longer than those of the second dorsal. Probably the specimen from which the Bloch-Schneider figure of R. glesne is taken was similarly damaged, as this figure (pl. xxiv., fig. 8) gives a small anterior dorsal fin supported by 7 rays all equal in length to those of the second dorsal. Bloch's semi-mythical Gymnetrus hawkinsii has a continuous dorsal with all the rays of about equal length.

In the Moeraki specimen the seventh and the ninth rays are perfect (pl. xxiii., fig. 1, r 7, r 9): the seventh is 1 foot 5.5 inches long, the ninth 1 foot 3.5 inches. Both taper gradually to the distal end and terminate in a somewhat irregularly lanceolate or spear-head shaped lobe about 1 inch long by rather less than half an inch wide (pl. xxiii., fig. 2). This lobe is formed by an expansion of the cutaneous covering of the bony ray, and is considerably thickened and corrugated on the surface. The bony ray also tapers gradually to its extremity and ends in a very fine point about ¼ inch from the end of the cutaneous lobe. The relative thickness of the spines corresponds very well with Hancock and Embleton's description quoted above: the first ray is about ⅛ inch thick at the base, the second to the fifth about 1/26 inch; the sixth and seventh a little more than ⅛ inch; from the eighth to the fifteenth the diameter gradually decreases. The first four, also, are closer together than the rest. The first five spines, although broken, have evidently not lost much of their length. The first is 1 foot 5.5 inches long, the second a little less, the third, fourth, and fifth about 13 inches; all the rest are broken off short. All the rays are connected by membrane with one another, and the last is similarly connected with the first ray of the so-called second dorsal, so that, as in some other specimens described, the two dorsals are continuous. The membrane uniting the first four rays extends to a distance of 9 inches from the base, and its edges do not appear to be perfect. In the 7th (perfect) ray (pl. xxiii., fig. 2) the terminal lobe is continued proximalwards by a thin membranous expansion along the posterior edge. The border of this expansion is evidently unbroken to a distance of 9 inches from the extremity, and I think it may be assumed that, for the rest of its extent, this membrane extended to the next ray. All the rays are quite smooth, presenting no trace of the spines or hooks described in some other specimens. My specimen would therefore seem to show that the rays of the crest, or so-called first dorsal fin, are united for about the proximal half of their length by membrane, that for the rest of their extent they are fringed posteriorly by membrane, and that they terminate in somewhat thickened lanceolate expansions. I have no means of determining whether all terminate in the same way, or whether, as in Cuvier's figure of R. gladius (pl. xxiv., fig. 6), some of them simply taper to a point. One would like, by the by, to know certainly whether Cuvier's figure is an exact representation of the specimen for which it was taken, or whether it is in any way “restored.” The extreme brittleness of the rays of the crest would seem to render the capture of an absolutely uninjured specimen extremely improbable; and it is therefore not unlikely that the figure in

question either gives the result of the examination of several specimens, or is restored more or less conjecturally. It would be interesting to know the precise history of the figure, but I have, unfortunately, no means of getting at it. In some species of Regalecus a caudal fin is present. Günther states that in R. russellii the caudal rays are distinct. In R. glesne he says, “it appears very doubtful whether the dorsal fin really was continuous with a caudal.” A distinct caudal fin continuous with the dorsal is, however, shown in the Schneider-Bloch figure of that species (pl. xxiv, fig. 9). The ordinary forked homocercal tail in Bloch's figure of Gymnetrus hawkinsii is certainly mythical. Cuvier's figure of R. gladius shows a delicate caudal fin (pl. xxiv, fig. 7), consisting of seven very fine rays standing out from the slightly enlarged extremity of the tail, and unconnected by any membrane. Of this fin Günther makes no mention in his systematic description of the species. In R. banksii and R. pacificus the tail ends in a bluntly-pointed extremity quite devoid of fin rays. In both there is a slight emargination on the ventral side, a short distance from the end. In all these respects the Moeraki specimen agrees exactly with the two last-named species. Lütken is of opinion that this absence of a caudal fin is due to “the peculiar mutilation or curtailment which the caudal extremity always seems to suffer to a greater or less extent in these fish.” I fail to see any evidence of such mutilation in my specimen. The number of rays in the pectoral fin is tolerably constant in the different species, the range being from 11 (R. banksii and R. russellii) to 14 (R. gladius and R. glesne): in my specimen there are 13 pectoral rays. In the present specimen the pectoral fin is remarkable for its vertical position, its line of attachment being almost perfectly horizontal. This appears to be the case also in R. gladius (pl. xxiv, fig. 6), and in R. pacificus (fig. 1). In R. banksii (figs. 3–5), and R. glesne (fig. 8), the fin has a markedly oblique position. The pelvic (ventral) fins are represented in all species each by a single long ray, the biradiate ventrals of Gymnetrus hawkinsii having been shown to owe their origin to an error on the part of the artist who drew the figure. The ventrals of R. russellii have also been erroneously described as biradiate. In the figures I have seen, the ventrals are represented as perfect in nos. 2 (pl. xxiv, fig. 6), 6 (fig. 4), 8 and 10 (fig. 8) of the list on pp. 285–6 above. In R. gladius the ventral is represented as terminating in an irregularly lanceolate cutaneous lobe, and as having a second somewhat triangular lobe on the post-axial side of the middle third of its length: no fringe of membrane is shown on either side of the ray. In the Banksian figure of the Nelwyn Quay specimen, (R. banksii? fig. 4), the terminal lobe is represented

as regularly oval and marked by radiating lines or ridges. In Schneider's R. glesne (fig. 8) the lobe is somewhat spatulate; neither in this nor in the preceding figure is there any fringing membrane. In R. gladius the length of ray is represented as nearly four times the height of the body, in R. glesne as fully three and a half times, in the Nelwyn Quay specimen as barely twice the height. In R. banksii, the pelvic rays are described by Hancock and Embleton as being fringed with membrane along the posterior (post-axial) edge: the same is shown in the Banksian figure of the Filey Bay specimen (fig. 5). Another specimen, probably of R. banksii, is described* Quoted by Hancock and Embleton, loc. cit., p, 17. as having ventrals three feet long and “fringed with a thin membrane on two sides.” No fringe is either mentioned or figured by von Haast in R. pacificus. In the Moeraki specimen the pelvic rays were both broken; the longer of the two was 37 inches in length, about 1/3 inch thick at its proximal end, tapering gradually to the fracture and fringed postaxially for a considerable distance by a delicate red membrane. In the form of the head the specimen under consideration resembles on the whole R. banksii; the “forehead” is, however, nearly straight instead of slightly concave as in that species and in R. gladius: in R. pacificus it is distinctly convex. There is also no difference of importance in the position and size of the eye, which, as in R. pacificus is slightly wider than high. The nostril (pl. xxiii., fig. 1, na) is apparently perfect on one side, and forms a single large oblique oval aperture, much larger and nearer to the eye than in R. banksii (fig. 3): in R. gladius (fig. 6) two widely separated nasal apertures are shown in Cuvier's figure. In Haast's figure no nostril is shown, but this is very probably owing to the coarseness of the photolithograph, in which most of the details of the head are completely lost. From the same cause the form of the operculum cannot be made out in Haast's large figure, but luckily the small outline figure shows it to have had a somewhat convex upper border produced into three points (pl. xxiv., fig. 1). The same is the case with R. gladius (fig. 6), with which, as with R. pacificus, my specimen agrees closely, differing mainly in the dorsal border of the operculum being, as a whole, less arched. As in R. gladius, too, the dorsal border of the opercular bone (op) is produced into two points, a third being formed by the anterior and a fourth by the posterior boundary of that border. The posterior boundary of the opercular is produced in the Moeraki specimen into a single point, which also marks the dorsal end of the sub-opercular: in R. gladius the latter bone apparently extends much further upwards. In R. banksii the opercular has an even border, its dorsal edge is somewhat concave and the sub-opercular is not indicated, having been mistaken by Hancock and Embleton for a branchiostegal.

The jaws have essentially the same characters as in other species, being extremely protrusible, and there are as usual six branchiostegal rays. The whole form of the head, including the mouth, in the figure of R. glesne, is probably incorrect, as it differs entirely from the most authentic figures. The specimen from which fig. 5 was taken had evidently had the jaws completely destroyed, and the artist has made the hinder boundary of the preopercular take the form of a mouth with a very remarkable grin. Regalecus gladius (and by implication R. telum) is said by Günther to possess small teeth: Yarrell makes the same statement of R. glesne. My specimen like all the remaining species of the genus is quite edentulous.* There is a strange discrepancy between Mr. Travers's account of the Nelson specimen, and the systematic description of R. gladius, copied in the “Catalogue of N.Z. Fishes” as applying to that specimen. R. gladius as the description sets forth, possesses teeth, whereas Mr. Travers distinctly states that in his specimen “the jaws appeared to be entirely destitute of teeth. In the Nelson specimen Mr. Travers states that “from the lower lip depended a large number of rigid slender barbules, about sixteen inches long and of a brilliant red colour.” No such structures appear to have been met with by any other observer. Is it possible that this part of the description was “compiled from information given” to the writer, as he himself says is the case with some portions of his account? None of the species of Regalecus possess scales, except on the lateral line; but the skin in R. banksii and R. pacificus is studded with numerous bony tubercles. In most of the other species it is raised into soft warts. Probably in all, also, the skin is covered externally with a delicate silvery coating, removed by the slightest friction. In R. gladius and R. telum the body is marked with greyish spots of somewhat less diameter than the eye. In R. banksii and R. pacificus there are instead irregular blackish wavy lines, more or less vertical in position, on the anterior part of the body. In both these species, as well as in R. glesne and R. grillii, the tubercles into which the skin is raised are arranged in four (three or four, R. glesne) longitudinal bands; or more correctly (in R. banksii and R. pacificus at any rate) the sides of the body are raised into four longitudinal ridges, cut off obliquely in front by the lateral line, and having the tubercles on them larger than in the intervening depressed bands. Von Haast's description of R. pacificus as resembling frosted silver seems to me more applicable to the present specimen than any of the terms used in describing other examples. The dark irregular bands had the same general disposition as in Hancock's and von Haast's specimens. When I first saw the fish, within twenty-four hours of its capture, they were very

indistinct, appearing like greyish clouds seen through the silvery coating. As the latter wore off, they appeared black and well-defined, the silvery coating at the same time assuming a tarnished appearance. The four longitudinal raised bands were well marked, and above the uppermost of them was a fifth less conspicuous band, becoming very indistinct in front. A transverse section of the body (pl. xxiii., fig. 3) shows that each of the five bands (r) corresponds in position to the line of attachment of one of the strong intermuscular septa, of which there are altogether six, the ventralmost corresponding in position to the lateral line (l.l.). If I understand Haast rightly, the lines of attachment of the intermuscular septa corresponded, in R. pacificus, to the depressed bands. The tubercles with which the skin is studded are hard to the touch, and I at first thought they were bony. As a matter of fact, however, they are composed of strong fibrous tissue, and consist, as shown by thin vertical sections, of pyramidal elevations of the dermis. The internal organs agree in every respect with Hancock and Embleton's description. The specimen was an adult female, but the ovaries contained no fully-formed eggs, the time of capture being evidently not near the breeding season. The muscles were white and firm when fresh; by no means “little coherent,” as stated by Günther. The vertebræ were roughly estimated by Hancock and Embleton at 110, in the present specimen they amount to 93. From the above description it will be seen that the Moeraki specimen of Regalecus differs (a.) From R. gladius (?), R. telum (?), R. glesne (?), and R. russellii in the absence of a caudal fin, (b.) From R. glesne (?), R. gladius, and R. telum in the absence of teeth, (c.) From R. pacificus in the great length of the anterior dorsal fin rays, in the absence of spines thereon, in the shape of the head, and in the fibrous nature of the dermal tubercles, (d.) From R. banksii in the shape of the operculum and in the non-ossification of the dermal tubercles, (e.) From all species in the number of fin rays and in the proportion between height and length. It is therefore impossible to assign it to any of the hitherto recorded species, and I propose to make it the type of a new species to be called R. argenteus. It must be borne in mind, however, that the analogy of Trachypterus lends strong support to Lütken's opinion that the proportions of the head and body, the number of fin-rays, and the characters of the tail differ greatly at different ages: this being the case the present species, like many of those previously established, must be looked upon as more or less provisional.

Regalecus argenteus, T. J. P. D. 15/190: P. 13: V. 1: Br. 6. Height of the body about one-tenth, length of the head about one-seventeenth of the total length. Eye one-sixth of the length of the head. Length and height of the head about equal. The fifteen anterior dorsal rays form a crest the height of which is more than double that of the head: its rays have their lower halves united by membrane, their upper halves having a narrow membranous fringe; three or four of them terminate in lanceolate cutaneous lobes, and they are not spinose. Ventral rays fringed posteriorly by membrane. No caudal fin. Four longitudinal ridges and an indistinct fifth extend from head to tail above the lateral line, by which they are obliquely cut in front. Surface studded with numerous hard but not bony tubercles which are largest and most elevated on the ridges; those forming the ventral edge are not perceptibly hooked backwards. Teeth absent. Silvery, with irregular wavy sub-vertical stripes and spots; forehead and membranous portion of snout blue-black; fins crimson. Measurements. Feet. Inches. Total length (jaws retracted) 12 6 Length of head (jaws retracted) 0 9 " " (jaws protruded) 0 11 Height of head (through centre of eye) 0 9 Height of body at posterior boundary of operculum 0 11 Height of body 2 ft. from head 1 2.5 " " 4 ft. " " 1 3.25 " " 5 ft. 6 in. from head (level of vent) 1 2.5 Height of body 4 ft. from tail 0 11.25 " " 2 ft. " " 0 9 Thickness of body 3 ft. 2 in. from head 0 3.5 " " 5 ft. 11 in. " " 0 3 " " 9 ft. " " 0 2 Diameter of iris 0 1.35 " pupil 0 0.5 Length of 1st dorsal ray (broken) 1 5.5 " 7th " " (perfect) 1 5.5 " 9th " " " 1 3.5 Height of 2nd dorsal fin 0 2.25 to 3 Length of pectoral fin 0 3 Base " " " 0 1.25 Length of ventral ray (broken) 3 1

Key to the Species of Regalecus. A. More than 250 dorsal rays. a. Teeth present: a caudal fin (?). α. Height = 1/19 length 1. R. gladius. β. " = 1/24 " 2. R. telum. b. Teeth absent: no caudal fin. α. Height = 1/19 length: D.12-15/264-290 3. R. banksii. β. " = 1/18 " D. 406 4. R. grillii. B. Fewer than 250 dorsal rays. a. Teeth present (?): a caudal fin (?). D. 8/120-8/120 5. R. glesne. b. Teeth absent. α. A caudal fin: D.4-5/224 6 R. russellii. β. No caudal fin. i. Height = 1/11 length: D.9/223 7. R. pacificus. ii. " = 1/10 " D.15/190 8. R. argenteus. Explanation of Plates XXIII. and XXIV. Plate XXIII. Fig. 1. Regalecus argenteus, side view of the head (¼ nat. size). br, branchiostegal rays: d. 21, first ray of 2nd dorsal fin: e, eye: fr, preorbital process of frontal bone: g, gills: is, isthmus: i.op, inter-opercular: lc, lachrymal bone: l.l, lateral line: mn, mandible: mx, maxilla: na, nostril: op, opercular: pc. f, pectoral fin: p.e, par-ethmoid (pre-frontal): p.mx, pre-maxilla: p.op, preopercular: pv. f, pelvic (ventral) fin: r. 7, seventh, and r. 9, ninth rays of crest: s.op, sub-opercular. Fig. 2. Regalecus argenteus, distal portion of 7th ray of crest (nat. size); x, point at which membrane became broken. Fig. 3. Regalecus argenteus, transverse section through the body, at about the middle (¼ nat. size). d.f, dorsal fin: g.c, gastric cæcum: h.a, hæmal arch, enclosing caudal artery and vein: l.l, lateral line: n.a, neural arch, enclosing spinal cord: r.r, tuberculated ridges: v, centrum of vertebra. Plate XXIV. Fig. 1. Head of R. pacificus after von Haast. 2 Tail " " 3. Head of R. banksii, after Hancock and Embleton. 4. " " (?), from Gray. 5. " " (?), " 6. " R. gladius after Cuvier. 7. Tail of " 8. Head of R. glesne from Yarrell, after Schneider. 9. Tail " "

Regalecus. T.I.P.ad.nat.del.