Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 56, 1926
This text is also available in PDF
(2 MB) Opens in new window
– 529 –

New or Rare Fishes of New Zealand.

[Read before the Wellington Philosophical Society, 28th October, 1924; received by Editor, 15th November, 1924; issued separately, 26th April, 1926.]

Plates 8792.

Family Carcharinidae.
Genus Prionace Cantor.

Prionace glaucum (Rondelet).

Galeus glaucus Rondelet, 1554, Pisc., p. 378.

Carcharias glaucus Cuvier, Reg. Anim., vol. 2, p. 126, 1817: Yarrell, Brit. Fishes, vol. 2, p. 381, 1836: Day, Brit. Fishes, vol. 2, p. 289, pl. 152, 1836: Guenther, Fische Suedsee, vol. 3, p. 479, 1910.

Carcharinus glaucus Jordan and Gilbert, Bull. 16 U.S Nat. Mus., p. 22, 1883.

Prionace glauca Jordan and Evermann, Bull. 47 U.S. Nat. Mus., p. 33, pl. 4, fig. 16, 1896; Bull. U.S. Fish. Com., vol. 23, p. 37, fig. 3, 1905: Snyder, Bull. U.S. Fish. Com., vol. 22, p. 515, 1904: Jordan and Fowler, Proc. U.S. Nat. Mus., p. 26, p. 613, 1903: Jordan, Tanaka, and Snyder, Jour. Roy. Coll. Sci. Tokyo, vol. 33, p. 13, fig. 5, 1913: McCulloch, Check-list Fish. N.S. Wales, p. 6, pl. 1, 1921: Waite, Rec. South Aust. Mus., vol. 2, p. 12, fig. 10 (glaucum), 1921: Phillipps, N.Z. Jour. Sci. & Tech., vol. 6, p. 262, fig. 5, 1924.

Galeus glaucus Garman, Mem. Mus. Comp. Zool. Harvard, vol. 36, p. 145, pl. 3, figs. 1–3, 1913.

A badly stuffed shark in the Dominion Museum, previously labelled “Carcharinus brachyurus, Lyall Bay, Wellington,” is referable to this species. Owing to the poorness of the example, my figure of the species (Phillipps, loc. cit., 1924) was not a good one. The mouth is not as there represented, and little idea of the long sweeping pectoral can be gained from the figure. Actually the pectoral is almost equal to the distance from its origin to tip of snout. Such is the case in no other shark, and this is apparently all that is required for its identification.

A further distinction, which possibly may not be so important, is that shown in the relative heights of dorsals in Prionace and Carcharinus, the two Australasian genera, so closely related on account of the absence of spiracles in both. For example, in the figure of P. glauca submitted by Jordan, Tanaka, and Snyder (loc. cit.) we find height of first dorsal to be over 7 ½ times length to c.p. Measuring from figures, we find all available representations of Prionace glauca have the relation of the dorsal height to length as above to be 7 to 7 ¾, while numerous figures of species of Carcharinus gave 5.3 to 6 for same relationship. Mr. Gilbert Archey found the relation of height of dorsal to length in a Canterbury Museum example from a New Zealand locality to be 7; but in the Dominion Museum specimen this relationship is over 7 ½.

– 530 –

Family Isuridae.

Genus Isurus Rafinesque.

Isurus glaucus (Mueller and Henle). (Plate 87.)

Oxyrhina glauca Mueller and Henle, Plagiost., p. 69, pl. 29. 1838: Schlegel, Jap. Pisces, p. 302, 1850: Dumeril, Elasm., p. 409, 1865.

Isuropsis glaucus Gill, Ann. N.Y. Lyc., vol. 7, p. 398, 409 (name), 1861: Jordan and Fowler, Proc. U.S. Nat. Mus., vol. 26, p. 623, 1903: Jordan, Tanaka, and Snyder, Jour. Roy. Coll Sci. Tokyo, vol. 33, p. 16, No. 38, 1913.

Lamna glauca Guenther, Cat. Fish. Brit. Mus., vol. 8, p. 391, 1870; Fische Suedsee, vol. 3, p. 484, 1910.

Lamna spallanzanii Day, Fishes of India, p. 722, pl. 186, fig. 2, 1878.

Isurus glaucus Jordan and Evermann, Bull. U.S. Fish. Com., vol. 23, pt. 1, p. 43, 1903: Garmen, Mem. Mus. Comp. Zool. Harvard, vol. 36, p. 38, 1913: McCulloch, Check-list Fish. N.S. Wales, p. 8, pl. 2, 1921: Waite, Rec. South Aust. Mus., vol. 2, p. 21, fig. 27, 1921: Phillipps, N.Z. Jour. Sci. & Tech., vol. 6, p. 268, 1924.

Body rounded, tapering anteriorly and posteriorly from under first dorsal. Head ¼ total length. Depth at pectoral is contained 7.61, or under first dorsal 6.19, in total length. Eye to snout 3.3 in head, which is 4 in total length. Snout to origin of dorsal is 2.47, and snout to origin of pectoral 4.13, in total length. Snout to first gill-opening equals distance from origin of ventral to origin of anal, or five times pre-oral length. Thirty-three myomeres on side of body between posterior margin of first dorsal and commencement of second dorsal. Gill-slits regularly decreasing in size posteriorly, the last being in advance of pectoral. First gill-opening advancing considerably towards its fellow on the ventral surface. Lateral line rises in a distinct ridge on caudal peduncle near dorsal surface. A more or less vertical ridge defines base of caudal, and a ridge at right angles to it separates caudal lobes.

Fins: Pectoral 1.3 in head, its base being 1.38 in eye to snout. Height or dorsal about 2 in head and origin nearer origin of pectoral than origin of ventral. Second dorsal slightly smaller than and rises in advance of anal. Upper caudal lobe more than ¼ lower caudal lobe; secondary lobe of upper caudal lobe 5.3 times upper caudal lobe. Lower caudal lobe 1.27 in upper caudal lobe.

Teeth: There are 4 teeth in front on each side large and conspicuous, while the third tooth on each side is considerably smaller. Jordan and Evermann (Bull. U.S. Nat. Mus., vol. 47, vol. 1, p. 48, 1896) have found similar conditions in Isurus dekayi (Gill) taken from Cape Cod to West Indies. As in the other American species, oxyrhynchus, our species has 13 to 14 rows of teeth. Teeth at sides more triangular, and those in front narrow, long, flexuous, and point either directly down throat or towards roof of mouth. No basal cusps.

Garmen (loc. cit.) recognizes six species of Isurus, one of which, Isurus glaucus, from Japan and the North Pacific, has recently been included in the New South Wales list by McCulloch (loc. cit.). Waite (loc. cit.) has included the same species from South Australia; in this we note several characters which would appear to distinguish it from Mueller and Henle's figure. From available descriptive matter it appears possible that the original figure they submitted was not a good drawing of the species; for certainly, if Mueller and Henle's figure is correct, both the New Zealand species and that figured from South Australia are new to science. This is improbable.

– 531 –

Mueller and Henle showed the relation of the lower to the upper lobe of caudal to be 1.44. In Waite's figure it is 1.36, and in the New Zealand examples 1.27. The New Zealand examples have origin of dorsal distinctly behind posterior pectoral margin, as is the case in the type, while Waite's figure shows the dorsal rising over hinder margin of pectoral. Such discrepencies may be due to faulty taxidermy where museum examples have been examined. A twist in the lower caudal fin to represent swimming motion might easily be responsible for an artist failing to show the comparative size of lower caudal lobe.

Isurus glaucus is perhaps one of the world's most famous sporting fishes. It abounds in considerable numbers around the northern part of New Zealand, and anglers from all parts of the globe annually journey to Bay of Islands to capture it. Its average weight appears to be between 220 lb. and 330 lb., and it is to be secured throughout the summer off most of our northern rocky coast-line. This species is the “mako” of the Maori, which name it still retains.

Family Galaxiidae.

Genus Galaxias Cuvier.

Galaxias burrowsius n. sp. (Plate 88.)

D. 11; C. 14 (true rays); A. 12; V. 5; P. 11. Br. 7.

Length of head is contained approximately 5 times in length to caudal peduncle, while depth of body is contained about 9 times in same. Diameter of eye is 1.7 in anterior margin of eye to snout. Lower jaw longest, maxillary extending to vertical under anterior margin of eye. About 51 myomeres visible on side of body.

Body not greatly compressed. Top of head and body somewhat flattened Behind nostrils on each side are a number of pores of two sizes. These are situated on top of head in neighbourhood of each eye, and encircle preopercular region to mandibles. Nostrils placed one on each side in front of eye at distance equal to diameter of eye. Eyes moderate. Two pyloric appendages.

Fins: Ventral atrophid with 5 rays. Caudal rounded. Pectoral extends back to ⅓ of distance from base to base of ventral; ventral reaches back less than ⅓ distance from its base to base of anal. Origin of dorsal far back above anal. Dorsal not triangular, but elongated as in Neochanna.

Teeth: Present only on premaxillaries, lower jaw, and tongue, there being distinct enlarged lateral canines in lower jaw, and 3 to 4 series of large curved teeth on tongue.

Colour: General ground-colour of body dark green, with darker markings on side which tend to arrange on lines of myomeres. A general yellowish tinge is present around most markings, and on head and tail. Caudal with its large procurrent attachment is of a deep-yellow colour marked with brown spots. Dorsal and anal fins are light brown, the pelvic or ventral white with red at base, and the pectoral brown. The operculum is green and under-surface of body white. Lateral line distinctly marked by means of a row of minute golden dots.

Type: In the Dominion Museum. Total length, 131.5 mm.

Affinities of Galaxias burrowsius: General shape of body and disposition of dorsal, caudal, and anal fins altogether show a strong superficial resemblance to Neochanna; but the presence of a ventral fin with 5 rays and dorsal with 11 rays indicates a close relationship to Galaxias. Specimens of Galaxias brevipinnis and Galaxias lynx in the Dominion Museum show a

– 532 –

like point of origin for the ventral. All members of the genus Galaxias so far recognized from New Zealand have 7 rays in the ventral, and in three species alone out of twenty-six recognized members of this genus (Regan, Proc. Zool. Soc. Lond., vol. 2, pt. 2, pp. 363–83, 1906) the number is reduced to 6. These are—one species, dissimilis, the type of which is in the Australian Museum, and also zebratus and punctifer, the only two species known from South Africa. In Galaxias burrowsius ventral has undergone a remarkable decrease in size and extends less than ⅓ the distance from base to origin of anal, while in all other New Zealand species of Galaxias ventral extends ½ to over ¾ the same distance.

Galaxias burrowsius is undoubtedly a degenerate form of Galaxias, and is to some extent a connecting-link between Galaxias and Neochanna. It is known only from one locality, West Oxford, in the South Island of New Zealand. Two examples were secured by Mr. A. Burrows and forwarded to the Dominion Museum. They were packed alive in a tin box together with a quantity of damp earth, sent by parcel-post on a journey lasting over thirty hours, and arrived alive and extremely active.

Of this species Mr. Burrows writes as follows: “A creek near my house is dry for a good part of the summer, but every spring it has small fish in it from 4 in. to 6 in. in length. They swim in small shoals, and I caught some and placed them in a waterhole. The waterhole became dry, and no fish were to be seen; but on digging the bank down I found in it holes shaped like a coconut in which the fish hid, each with a small entrance. The holes were very smooth inside, and could hold water for a long time if the entrance was carefully closed from the inside.”

Family Gadidae.

Genus Lotella Kaup.

Lotella rhacinus (Forster). (Plate 89.)

Gadum rhacinum Forster, in Bloch and Schneider, Syst. Ichth., p. 56, 1801.

Lota rhacina Richardson, in Dieftenbach, Travels in N.Z., vol. 2, p. 222, 1843.

Gadus rhacinus Forster, Descr. Anim., p. 304, 1844.

Lotella rhacinus Guenther, Cat. Fish. Brit. Mus., vol. 4, p. 347, 1862; Hector, Cat. Col. Mus., p. 80, 1870; Hutton, Cat. N.Z. Fish., p. 46, 1872; Hector, in Hutton, Cat. Fish. N.Z. p. 116, 1872; Hutton, Trans. N.Z. Inst., vol. 5, p. 266, 1873; ibid., Trans. N.Z. Inst., vol. 6, opp. p. 104, pl. 18, fig. 74, 1874; ibid., Trans. N.Z. Inst., vol. 22, p. 282, 1890.

Physiculus rhacinus Hutton, Index Faunae N.Z., p. 48, 1904: Waite, Rec. Cant. Mus., vol. 1, p. 18, 1907; ibid., Rec. Cant. Mus., vol. 1, p. 318, 1912.

D. 5 + 58; C. 26; A. 52; V. 2 + 7; P. 23. Br. 7. L. lat. 120.

Body stout and elongate, compressed laterally; compression being greatest from second dorsal to anal and least beneath first dorsal in region of pectoral and immediately below it. Depth at vent 3.4, and length of head nearly 4 in length to caudal peduncle.

Head: Interorbital space raised just above eye on each side, with a tendency to become concave on top in specimen examined. Eye large, deeply set, its diameter being contained 5 times in length of head. Upper jaw longer; maxillary extends backwards to vertical under pupil of eye. Mandibular barbel greater than ½ length of pectoral.

Teeth: 15 large pointed conical teeth irregularly placed on outer margin of each side of lower jaw; 12 teeth on outer margin of each side of upper jaw, similar save that they are smaller amd more irregular. Villiform band of teeth immediately inside outer row in upper jaw, while small scattered teeth in bands occupy the same position in lower jaw. Oval patch of 12 rows of small teeth present on head of vomer.

Picture icon

Isurus glaucus (Mueller and Henle).

Picture icon

Galaxias burrowsius n. sp.

Picture icon

Lotella rhacinus Forster.

Picture icon

Aotea acus n. sp.

Picture icon

Xiphias gladius Linn.

Picture icon

Istiophorus gladius Broussonet.

– 533 –

Fins: The first dorsal commences over origin of pectoral, being placed relatively farther back than in Australian species. Ventral commences a little farther forward in reference to pectoral than in Australian species. Second outer filamentous ventral ray reaches nearly ⅔ distance from base of ventral to vent. Vent situated slightly less than ⅓ of total length of fish from snout to end of tail. Anal commences 10 mm. behind vent. Caudal small and rounded.

Colour: Uniform dark red. A dark line runs from upper margin of operculum to dorsal surface. Other dark markings, more or less distinct, are apparent on head, running parallel to branchiostegal rays. Below fins colour lightens to a pale reddish-yellow. Fins uniformly dark.

The following key will serve to differentiate the New Zealand and Australian species of Lotella:—

Depth at vent 3.4 and length of head nearly 4 in length to caudal peduncle rhacinus.

Depth at vent nearly 4 and length of head 4.5 or over in length to caudal peduncle callarias.

Callarias has a longer maxillary bone than rhacinus.

Hutton (loc. cit.) called this species “hake,” a term used to refer to Merluccius vulgaris in the North Sea. If we are to continue to apply English names to like New Zealand species, Merluccius gayi has prior right to the name “hake”; but, unfortunately, Jordanidia solandri is in Blenheim, Nelson, Picton, Wellington, and Napier caught and sold as “hake.” This I have verified.

Lotella rhacinus is not uncommon in Cloudy Bay, and during 1924 small numbers were sold in Wellington as “Cloudy Bay cod” and “rock-cod.” Working on the principle of applying English names to like and related species in our waters, Lotella rhacinus might rightly be described as a haddock or a cod. The first specimen which came under my notice was secured by the late Dr. Moorhouse (of Christchurch) off Cape Brett, and presented to the Dominion Museum. This example is here figured, and is stouter than most Wellington specimens.

The following key will serve to differentiate the two allied genera, Physiculus and Lotella:

A row of strong large teeth with bands of small ones; scales very small Lotella.

Teeth subequal in size, in bands, scales large Physiculus.

I have to thank Mr. A. R. McCulloch, of the Australian Museum, for comparing a specimen from New Zealand with the Australian examples.

Family Aoteaidae new:

Aotea n. gen.

Body naked, extremely elongate, cylindrical, not appreciably compressed except in caudal region, tapering only in posterior quarter of length. Body and head covered with a semitransparent skin, which peels off on being handled. Lips well developed; snout acute and compressed, overlapping mandible, mouth being underneath. Lateral line present. Fins absent. A semitransparent portion of back towards tail may represent the dorsal fold of certain Synbranchidae, or may be caused by shrinkage. Minute teeth present in several series on palatines and vomer extending to inter-maxillaries. Mandibular teeth in single series. All teeth minute and approximately the same size, though those on symphysis appear stronger.

– 534 –

Aotea acus n. sp. (Plate 90.)

Length of head approximately 24 and depth behind head 50 in total length. Eye 12 in length of head. A minute posterior nostril situated on each side of head about diameter of eye in front of posterior margin in line with centre of pupil. Upper maxilla overlaps mandible by a distance equal to diameter of eye. Angle of jaws reaches back to distinctly behind posterior margin of eye. Eye to snout equals depth of body behind head. Snout not anteriorly rounded as in Synbranchus.

Owing to abrasions and swellings, apparently caused by parasites, the anus has been obliterated. Species is unfortunately not in a fit state for dissection, but position of vent appears to be under one of the numerous swellings on the ventral surface in anterior portion of body. A hard folded portion beneath body posterior to head apparently indicates gill-openings, which are similarly placed in Synbranchus. Along parts of anterior body-wall segmentation is clearly visible.

Affinities of Aotea acus: Aotea acus differs from all other genera of the suborder Synbranchiformes in the possession of vomerine teeth, and also, as far as can be learned from available literature, in the extremely elongate body. Guenther (Cat. Fish. Brit. Mus., vol. 8, p. 12, 1870) has divided what was then regarded as the family Symbranchidae * into the following three groups—(a) Amphipnoina, (b) Symbranchina, (c) Chilobranchina—of which the first two have vent in posterior half of length, while the last has vent in anterior half; but all Chilobranchina known to Guenther were without vomerine teeth.

Weber and Beaufort (Fish. Indo-Aust. Archip., vol. 3, p. 411, 1916) discuss the order Synbranchoidea, and give a key to the three genera of Synbranchoidea. It would appear that Aotea acus shows certain affinities to Monopterus albus, common in the fresh waters of Sumatra, Borneo, China, and Japan; but its serpent-like head, attenuate body, and vomerine teeth indicate otherwise. The New Zealand species appears to show considerable differentiation of arrangement of frontal, parietal, and supra-occipital bones of the skull; but without injury to the type it is impossible to determine how far bones have been modified.

Regarding the Synbranchiformes, Goodrich (Treatise on Zoology, vol. 9, p. 408, 1909) writes: “A small group of very highly specialized fish whose affinities cannot yet be determined. They have a superficial resemblance to the eels, from which they differ in many important osteological characters, and in the possession of closed ovisacs. The air-bladder is absent. The skull is like that of the Clupeiformes; the parietals meet, but the maxillae are almost excluded from the margin of the mouth. The group is unknown in a fossil state.”

Aotea acus is related to the order Carenchelyi, or long-necked eels, in its slender body and serpent head (Jordan and Evermann, Bull. U.S. Nat. Mus., vol. 47, p. 343, 1896). This order is represented by a single species taken in the Gulf Stream at 1,022 fathoms; but it differs from the New Zealand species chiefly in the presence of fins. The Synbranchiformes are represented in Australia by the family Cheilobranchidae, of which two small species, 2 in. to 4 in. long, are common in coastal rock-pools and inshore waters (McCulloch, Check-list Fish. N.S.W., p. 22, Nos. 76A and 76B, 1920).

Type: The type was taken from the stomach of a snapper (Pagrosomus auratus) secured in Cook Strait, and forwarded to the Dominion Museum by Mr. H. P. Washbourn, Nelson. It is assumed that the species is marine,

[Footnote] * Generally wrongly spelt “Symbranchidae” instead of “Synbranchidae.”

– 535 –

and it is possibly an inhabitant of deep water which had died or left its abode owing to disease.

Family Girellidae.

Genus Girella Gray.

Girella cyanea Macleay.

Girella cyanea Macleay, Proc. Linn. Soc. N.S. Wales, vol. 5, p. 409, 1881: Ogilby, Proc. Zool. Soc. Lond., p. 394, 1887: Waite, Rec. Aust. Mus., vol. 5, p. 167, pl. 20, 1904: Roughley, Fish. Australia, p. 54, pl. 13, 1916: Phillipps, N.Z. Jour. Sci. & Tech., vol. 4, p. 116, 1921.

Hitherto known from waters of New South Wales, the Kermadec and Lord Howe Islands, I have recently recorded Girella cyanea from Hauraki Gulf, where it appears to be caught only by the Maori. It is not common, and is rarely seen south of Whangarei. Structurally it resembles Girella tricuspidata, but differs chiefly in having a deeper body, smaller scales, and outer teeth in single rows. Messrs. W. R. B. Oliver and H. Hamilton, of the Dominion Museum, have recently taken G. cyanea off Poor Knights Islands.

Family Agrostichthyidae.

Genus Agrostichthys Phillipps.

Agrostichthys parkeri (Benham).

Regalecus parkeri Benham, Trans. N.Z. Inst., vol. 36, p. 198, pl. 9, 1904. Phillipps, N.Z. Jour. Sci. & Tech., vol. 6, p. 232, 1923.

Agrostichthys parkeri Phillipps, Proc. Zool. Soc. Lond., pt. 2, p. 540, figs. 1, 2, 1924.

This species, which I have made the type of a new genus, is of peculiar interest in that its general outline of body agrees with the Regalecidae, while the shape of the head and disposition of the teeth show considerable affinity to the Trachypteridae. It differs from each, however, in the enormous extent to which the body of the adult is elongated, and also in the long maxillary plate.

Suborder Xiphiiformes.

Ventral fins absent; mandible less than ½ maxillary Fam. Xiphiidae.

Ventral fins present; mandible ½ or more than ½ maxillary Fam. Istiophoridae.

Family Xiphiidae.

Genus Xiphias Linnaeus.

Xiphias gladius Linnaeus. (Plate 91.)

Xiphias gladius Linnaeus, Syst. Nat., ed. 10, p. 248, 1758: Guenther, Cat. Fish. Brit. Mus., p. 511, 1860: Hutton, Trans. N.Z. Inst., vol. 22, p. 278, 1890: Jordan and Evermann, Bull. U.S. Nat. Mus., vol. 47, pt. 1, p. 894, 1896; Bull. U.S. Fish. Com., vol. 23, p. 168, fig. 61, 1903: Hutton, Index Faunae N.Z., p. 43, 1904: Waite, Rec. Cant. Mus., vol. 1, p. 25, 1907.

Ziphius gladius Hector, Trans. N.Z. Inst., vol. 7, p. 246, 1875.

Ziphias gladius Cheeseman, Trans. N.Z. Inst., vol. 8, p. 219, 1876.

Xiphius gladius Sherrin, Fish. N.Z., p. 96, 1886.

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

D. 2 + 14/3; P. 1 + 12; A. 1 + 9/3. Br. 7.

Head approximately 5, depth at opercular edge excluding dorsal sheath nearly 6, and depth at anterior anal 7.73 in length from tip of mandible to tip of caudal. Tip of spear to eye approximately equal to distance between

– 536 –

origin of pectoral and first anal. Mandible pointed and ½ length of head. Eye 6.08 in head. No teeth. Body generally compressed laterally, naked, more or less rough, and compressed dorso-ventrally in region of caudal peduncle. An enlarged keel on each side about middle line on caudal peduncle, which is broader than deep.

Fins: First dorsal commences above gill-openings, falcate, and elevated, rising to a height almost equal to length of head. Second dorsal small, rising well back on dorsal surface and reaching over lateral keel. Length of pectoral approximately equal to depth at opercular margin. First anal sometimes originates at a point a little nearer operculum than tip of caudal. Second anal is in front of second dorsal and about the same size. Ventrals absent. Caudal forked.

Colour: “Upper surface has a general greyish-purple tinge merging to dark grey on dorsal surface. Fins of uniform greyish-green. Under-surface white” (Clarke's drawing). Jordan and Evermann describe the Hawaiian example as follows: “Colour dark metallic-purplish above, dusky below, sword almost black above, below lighter; fins dark, with silvery-sheen.”

Clarke's drawing of the species, here published for the first time, represents a swordfish having a total length of 11 ft. 9 ¾ in., washed ashore on Hokitika beach. It appears to agree very well in all essential details with Cuvier's figure of the species, reproduced by Jordan and Evermann (1903, loc. cit.). The large fossa shown on the sides of Clarke's example are not shown in figures of this species from the Atlantic, and do not appear in a large stuffed example among the collection of New Zealand fishes in the Dominion Museum. The young of Xiphias gladius is known to be possessed of teeth and two connected dorsal fins. Though not common in the Pacific, the species is highly valued as food in southern Europe, and is the object of extensive fisheries in the Atlantic.

Family Istiophoridae.

Genus Istiophorus Lacepede.

Istiophorus gladius (Broussonet). (Plate 92.)

Scomber gladius Broussonet, Mem. Acad. Sci., p. 454, pl. 10, 1786.

Istiophorus gladifer Lacepede, Hist. Nat. Poiss., vol. 3, p. 374, 1802.

Histiophorus indicus Cuvier and Valenciennes, Hist. Nat. Poiss., vol. 8, p. 293, pl. 229, 1831: Valenciennes, Regne Anim. Illustr. Poiss., vol. 49, p. 124, pl. 53, fig. 1, 1836.

Histiophorus sp., Knox, Trans. N.Z. Inst., vol. 2, p. 13, pl. 1, 1870.

Histiophorus herschelhi Hutton, Trans. N.Z. Inst., vol. 8, p. 216, 1876.

Histiophorus gladius Guenther, Brit. Mus. Cat. Fish., vol. 2, p. 513, 1860: Day, Fish. India, pt. 2, p. 198, 1876: Castelnau, Proc. Linn. Soc. N.S. Wales, vol. 3, p. 352, 1879: Macleay, Proc. Linn. Soc. N.S. Wales, vol. 5, p. 522, 1881: Goode, Rep. U.S. Fish. Com., 1880, p. 309, pl. 8, 1883: Ogilby, Cat. Fish. N.S. Wales, p. 25, 1886.

Histiophorus herschellii Hutton, Cat. Fish. N.Z., p. 14, 1872; Trans. N.Z. Inst, vol 22, p. 278, 1890; Index Faunae N Z., p. 43, 1904: G. M. Thomson, Hist. Portobello Fish-hatch., p. 79, 1921 (herschelli).

Tetrapturus indicus Waite, Mem N.S. Wales Nat. Club, p. 42, 1904; Rec. Cant. Mus., vol. 1, p. 25, 1907.

Istiophorus gladius McCulloch, Rec. Aust. Mus., vol. 13, pt. 4, p. 137, pl 24, 1921.

D. 47 + 7; A. 9 + 7; P. 17; V. 2. Br. 7.

Depth under third anterior dorsal ray (excluding dorsal sheath) 1.7 in length of head measured from tip of mandible to opercular margin, or almost

– 537 –

exactly 6 times in length from tip of mandible to highest point on lateral keels at base of caudal. Eye 10 in head. Eye to tip of rostrum approximately equal to length of head. Rostrum straight, broader than deep, covered exteriorly by rough rounded spinulex which continue posteriorly well toward angle of jaw. Mandible pointed in front and extending backwards beyond posterior margin of eye. Nostrils small, situated in front of eyes.

Body compressed, broadest in region of pectoral and thence narrowing backwards. A broad fleshy fold on each side of body on dorsal surface extending backward to commencement of second dorsal, thus forming a sheath to accommodate first dorsal rays. Ventral and anterior anal rays have similar sheaths. Bases of second dorsal and second anal uncovered. Body covered entirely with pointed scales, shortest on dorsal surface and longest on ventral. Caudal peduncle deeper than broad, with two keels on each side forming distinct ridge.

Pectoral relatively longer than in McCulloch's example, reaching from base more than ½ distance to origin of anal, or being longer than depth under anterior dorsal rays. Dorsal commences above operculum, having first two rays short and third longest. Thence rays decrease in size to second dorsal, which consists of 7 rays, the first being longest and others decreasing posteriorly. Caudal slender and much forked, with two raised keels on each side of base. Towards posterior margin of the strong base of caudal lobes are 3 ridges which may be supplementary keels.

Origin of first anal to origin of second anal equals height of dorsal above opercular margin. Second anal is below second dorsal, its first ray being longest and others smaller posteriorly. First anal has third ray longest. Ventral commences a little behind origin of pectoral, and consists of 2 fused rays forming a rod on each side extending backwards as far as pectoral.

Colour: Upper surface dark brown with several curious vertical bands of white, some of which extend from dorsal surface to lateral line. This ground-colour of brownish-black extends forwards to eye and on to rostrum in front and sharply defines the greater part of upper dorsal surface, being posteriorly confined to upper portion of caudal peduncle. Ventral surface white, in marked contrast to dorsal. All fins of a dull reddish-brown. Anterior dorsal is not in this case uniformly black as in McCulloch's description of the young.

Discussion: This species is the common swordfish of Bay of Islands. It is here described from a specimen washed up at Napier and purchased from Messrs. Boyd during 1921. A cast is in the Dominion Museum. As a sporting fish of our northern waters this species has recently risen to fame, and, together with the mako and kingfish, is a favourite with visiting anglers. The young of this species, so well described by McCulloch (loc. cit.), apparently differs in two important respects from the New Zealand species—(1) the shorter pectoral, and (2) the incongruous length of the dorsal fin. Of these differences the former is the more important. Possibly much of the dorsal breaks away as the fish grows up; and it can only be assumed that, as the dorsal disappears, the pectoral increases in length and becomes in the adult the long appendage shown in the accompanying plate.