Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 31, 1898
This text is also available in PDF
(1 MB) Opens in new window
– 78 –

Art. VIII.—Notes on New Zealand Galaxidæ, mare especially those of the Western Slopes: with Descriptions of New Species, &c.

[Read before the Wellington Philosophical Society, 14th March, 1899.]

Plates IV. and V.

Though the common eels are almost the most ubiquitous of our fresh-water fishes (in spite of statements made that in some streams the prevalence of waterfalls prevents their passage to the upper waters), due merit must also be given to the equal or greater prevalence of more or less varieties of the Galaxidæ. As our Islands are famed for remains of varieties of struthious birds, so we should gather ichthyo-logical fame for the great number of varieties of these fishes, and it would be most interesting if evidence should be obtained of their geological existence also. It is a most engaging family, not only on account of its affinities with or resemblance to the Salmonidæ and Esocidæ, but also on account of its known extended habitat in the Southern Hemisphere—Western Australia, South Australia, Victoria, Queensland, Tasmania, New Zealand, Chatham Islands, Chili, Patagonia, Tierra del Fuego, and Falkland Islands.

The family certainly arrives at its greatest known development in our Islands, and its least—as far as has been described at present—in the southern parts of South America.

One of the members of the family—G. attenuatus—periodically descends to the sea in January, February, and March, where it spawns, returning in March, April, and May. The young begin to make their appearance in the rivers sometimes as early as the end of June, but they definitely commence to arrive in August, the shoals increasing in size and number in September and October, and keep practically unmixed with other fry until the end of October. At this time, and in November, the shoals begin and continue to consist of a mixed character—fry of Eleotris, Retropinnæ, and Prototroctes forming a greater percentage of them till the in run ends. The young fry of Eleotris in November are thiekly swarming. They and the fry of the Retropinnæ evidently follow the fry of G. attenuatus until they are large enough to prey on them. Frequently you will see small or tiny Retro-pinnæ or Eleotris with a half-gorged whitebait sticking out of their mouths almost equalling in size the swallowers.

– 79 –

The extent of the shoals of the whitebait in the South Island west coast rivers at times was incredible; often I have seen the surface of the Chinamen's gardens at the Buller, Grey, Teremakau, Hokitika, &c., for several acres each in extent, covered some inches in depth with these fry, used as top-dressing manure. This was when an expressly heavy run occurred, making it non-practicable for our Celestial migrants to dispose of them at a price which they considered paid for the labour of hawking; or when the shoals were too greatly mixed with “cock-a-bullies”; or when they were too lazy to dry and export them to China, which they sometimes did to a considerable extent, especially when the more periodical large shoals of smelts were running. These they carefully dried and packed in strong “upper” leather bags for export, obtaining from 3s. 6d. to 5s. per pound for the dried article from the middleman or “boss,” who probably reaped a much greater price still in China. This wholesale destruction for manure purposes, I am happy to say, was of late years stopped by preventing the use of the long shingle abutments with the set nets at intervals. It was piteous at those times to see the enormous quantities of young grayling (upokororo) which were destroyed.

Now that the whitebait (Galaxias attenuatus fry) is being utilised for canning—which is beginning to form a considerable industry—one hopes that the supply will last, and be properly fostered, to allow sufficient to be left for annual reproduction. But a safeguard in this is that it will be impossible to collect from all the rivers the supplies required for the canneries, and I expect it will be some little time before they are established on any but a few of the larger ones.

The Maoris of the West Coast, South Island, called the young fry “inanga” and the adult fish the “mai-tai,” or “mahitahi,” or, as they described it, “te whaea o te inanga.”

At the time of the advent of the fry I have frequently, and at several places, seen large shoals of the “inanga” at sea, and have caught specimens in verification, and have constantly observed them washed up by the breakers on to the beaches near the mouths of the large rivers, evidently when skirting the coast to enter them, and I have seen them dragged on shore in estuaries by the ground-ropes of the seine. A spate in the whitebait rivers will prevent the inrun until the fresh is over.

The early-run fish keep their immaculate colour much longer in the fresh water than those which come in towards the end of the season, these last, in fact, sometimes beginning to darken after a few hours' sojourn in the fresh element.

The Europeans of the West Coast, South Island, call the adult fish “cents” or “scents,” whilst the youth of Taranaki

– 80 –

District designate them “mingies”—I suppose an evolution from “minnows,” by which name they are commonly known in many other parts of New Zealand.

The whole of the family are very voracious fish, the most meek- and harmless-looking attenuatus being proportionately the worst in that respect. When turning out trout-fry (by necessity, not having developing-ponds therefor) at that supremely ridiculous age which by absurd fad had been adopted by so many for their liberation—I allude to the period soon after absorption of the umbilical sac—I have often seen three or four of these ubiquitous little rascals rush into the shoal—from goodness knows where—even when the greatest precautions had been taken in selection of a site comparatively free from them—and with lightning-like darts here and there avail themselves of the expensive change of diet so easily afforded, and completely decimate them in a very few minutes.

In captivity, both in the pond and aquarium, experiments ranging over many years have shown that G. attenuatus must return to salt water, otherwise it dies. A certain proportion each year have to take their departure to the ocean or be provided with salt water if their continuation in life is required.

G. attenuatus runs up to the heads of all streams, of whatever size and almost of whatever nature. Even the most trivial of little “trickles” of water will be found to be inhabited by them. Dark or clear, thin or thick water—none comes amiss.

They very seldom exceed a total adult length of 6 in. The largest ever observed by me in either of the other colonies or New Zealand in an experience of forty-four years was barely 7 in. in length. Average length of adult fish, 4 ½ in. to 5 in. (total). Attenuatus will take a fly, in brisk or still water; and in its adult state is a sweet little fish gastronomically.

As this species is ocean-frequenting, its greater extension of habitat is not so much to be wondered at, but it is difficult for one to surmise whether this capacity is an evidence of enhanced development or not. Certainly, in its corporeal characteristics, as compared with the more giant types we possess, and which I refer to hereafter, it has in my idea depreciated in type.

So far as my researches have gone in the streams of the North Island, the place of the five larger varieties of the family (as enumerated by Professor Hutton*) frequenting those of the South Island are occupied by G. fasciatus and G. abbreviatus only. Numerous native names are given by several authorities for varieties. These might represent two or three definite species, but more probably only mean the local native

[Footnote] *Trans. N.Z. Inst., vol. xxviii., p. 317.

– 81 –

names given to the two different species I mention at various stages of their growth, in accordance with their usual system of nomenclature, by which a human being might be “Brown” when a boy, “Smith” as a hobbledehoy, or become “Robinson” in old age. My own observations are largely strengthened by inquiries amongst our survey class of bush-wanderers, who have an unfailing aptitude for discovery of anything in the edible variety on legs, wings, or using fins. My H. M. Skeet is the only one of the latter who has informed me he has found a spotted variety, but I still live in hope of finding the existence of some of our southern types, especially as Neochanna are common to the two Islands.

The larger Galaxias of the Taranaki streams, though coloured like fasciatus and answering in other respects to the description of that species, certainly, in those I have examined, does not, in position of ventral and anal fins, agree with the arrangement of Professor Hutton before quoted; and, in addition to the three other varieties I describe at length, I have also given a new description of this fish, with a drawing for comparison (Plate V.).

With due deference to Professor Hutton's division of the New Zealand species before referred to, I cannot quite get all the types I have found in Westland to correspond with his divisions as extant; and, viewing his close attention and large scope of observation in matters scientific, &c., I am of the idea that two of the Westland types may not have come under his ken, and am positive that, one has not. The last one of these is an interesting one, as more resembling Neochanna in form, “if the ventrals were wanting,” than any of the hitherto described Galaxidæ.

For some years after 1870—the date of my arrival in New Zealand—at occasional intervals I obtained a few isolated specimens of this fish amongst fry of G. attenuatus, Retropinna richardsonii, Eleotris gobioides, and Prototroctes oxy-rhynchus, the heterogenous collection making up the so-called “whitebait” of the end of the inanga season in the Westland rivers, though, strange to say, in my long experience and ceaseless watching, season after season, I never found amongst this collection the fry of any of the larger indigenous Westland Galaxidæ. It was not until 1887 that I got an adult fish, one being caught by Mr. H. L. Robinson when fishing for grayling (upokororo) in the Hokitika River, just above Glossop's Ferry. Since that date I have taken them—but very occasionally—in the Kanieri River and Lake, Kawhaka Creek, Frosty and South Creeks, and the tributaries of the last mentioned, which can best be depended upon for the production of a specimen, as they are not by any means common. I have designated this variety G. robinsonii.

– 82 –

Another of the Westland species I refer to, which, though agreeing perhaps in some measure with alepidotus, in greater particular's differs. These differences, with all the varieties I now deal with, I have—-by description and tabular arrangements prepared on even system—rendered easier for inspection and comparison. I give two drawings of this species also, to show the distinction between the more moderate-sized and larger fish; but the principal points of comparison remain constant (Plate IV.). This variety grows to such constant large size that it would have been more appropriately called grandis, if the late Professor von Haast had not already adopted such specific name for the fish Professor Hutton now classifies as a variety of G. brevipinnis. However, as the South Island west coast Maoris always call this species “kokopu,” and as such name has not yet been specifically used, I propose to distinguish it as Galaxias kokopu.

G. kokopu is of excessively slow growth as compared with the rest of the family, and prefers to frequent those smaller rivers and streams which have rocky or hard gravelly beds. It is sometimes caught in the larger snow-fed rivers, but only occurs there—in my experience—near the mouths of tributaries with the features noted. It evidently does not like too great a continuous current to exist in, as is generally the case with the Westland wholly snow- or snow- and rain-fed, large, open, shingle-bedded rivers—practically huge mountain torrents—as you occasionally find, after spates, a specimen or two washed up on the beaches at the mouths of such rivers as have suitable tributaries near their embouchure to the ocean.

Professor Haast writes, in his article on G. grandis* “This species occurs also at the West Coast, where I obtained it in Lake Hall [Paringa Lake], the outlet of which falls into the Paringa River.” This shows he must have observed this West Coast variety I now refer to without having properly defined it at the time, as he obtained and described his grandis some years after his visit to South Westland; also, G. kokopu is commonly obtained in the streams about the Paringa Valley, and north and south of it, of which more anon.

Kokopu differs from grandis in very many points, nothing but the colour—apart from the other general family characteristics—being deducible from the description as at all corresponding; and this colour, though described as “yellowish spots and short streaks,” is noted as “on the back and head they are small and of rare occurrence,” which is quite different from G. kokopu. Neither, as before mentioned, does it agree with the principal points of alepidotus, as defined by

[Footnote] *Trans. N.Z. Inst., vol. v., p. 278.

– 83 –

Günther (Cat. Fishes, vol. vi., p. 208), by Captain Hutton's “Catalogue of Fishes of New Zealand” (p. 58), or Professor Hutton's divisions as before mentioned; nor with brocchus, which Günther (ib., p. 210) defines as a variety fasciatus and Professor Hutton as one of alepidotus. I have captured them year by year in Westland—from 1870 to 1894—-of all sizes, from a couple of inches to a maximum total length of 23 in. and weight of 6lb., and in all they keep their robust proportions, though the very large ones appear to have a finer head, on account of the greater development of the “corporation.” They take the fly very fairly, especially if it is presented in the form of a large live moth or blow-fly, and in captivity in the aquarium make very handsome and docile pets, soon learning to put their heads out of the water to take a blow-fly from the fingers. As with the English brown-trout, the larger fish are perfect tyrants, and keep the whole of a large pool under subjection. After catching the very large ones, of which a pool will only contain three or four, the next in size, but of larger number, become apparent, and take possession. These, when caught, are represented by a still smaller but more numerous draft. I presume these come out of the very numerous nooks and crannies and sheltering-places about small and large snags, where they are comparatively safe from the pursuit of the large tyrants. To enable these to appear again the pool must be left in abeyance for two or three seasons. The spots, crescents, and old Arabic-script-like markings of this fish are much larger in proportion in the smaller fish, but always keep of the same general figure; and this species I have found, like abbreviatus, proves most constantly free from aberration.

In the time of the notorious “Hunt's Bush” to South Westland it was a godsend to the hungry and fast-travelling prospectors, who, on following up the rocky-pool streams running down the sides of the lower ranges abutting on the sea-coast, obtained easy and welcome supplies of the large ones, thus affording, as several told me, “a good galvanisediron bucket full of solid-fleshed food for the evening meal; and a breakfast as well.” Their flesh is white and sweet, and very good if properly cooked, but, being so solid, requires quite double the time of cooking most fish.

In the very hot and dry late summer and autumn seasons—which you get to perfection in the usually much-maligned West Coast, South Island, climate—when the pools in the creeks and streams which it inhabits are much curtailed thereby, and get a “bloom” on the surface (in miniature like the “Indian summer” of North America), this variety is occasionally infested with a long thin red flesh-worm, which cysts up in the flanks, and also lies in the thick muscles along

– 84 –

the back. These, when the water improves, eat their way bodily out through the exterior.

Kokopu feeds a great deal on the fresh-water lobster, and has many of the habits of the Home trout—inhabiting the deeper streams, lying close to the surface at times, sometimes sheltering under a patch of foam at the foot of a miniature cascade or fall, and rising to the surface in the evening and taking the flies or moths.

The third Westland species I enlarge upon, more frequently inhabits the sluggish and muddy-bottomed creeks, but is also found in company with G. kokopu in the gravel-bottomed and some of the rocky creeks. In its proportions it somewhat approximates to the description of fasciatus, though it grows much larger, but seldom beyond 10 in. in length. It is not as hardy in the aquarium as G. kokopu, and has generally the same feeding habits, except that it does not take a surface-bait as well. Strange to say, it is seldom, if ever, troubled with the flesh-worms before mentioned. I have distinguished this one with the specific name of postvectis, on account of its peculiar and constant marking.

Order Physostomi.

Family Galaxidæ.
Genus Galaxias.

Fin-ray Formulæ and Branchiostegous Rays.

D. P. V. A. C. B.R.
Galaxias kokopu 14 13–14 7 15–16 16 8
" postvectis 11 14 7 15 16 8
" robinsonii 13 15 7 15 16 8
" fasciatus 10 13 7 13 16 9
Vertebræ. Head is to Total Length.
Galaxias kokopu 35 + 26 = 61 As 3 is to 13
" postvectis 35 + 25 = 60 As 5 is to 23
" robinsonii 35 + 25 = 60 As 7 is to 41
" fasciatus (not counted) As 3 is to 16

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

G. kokopu (larger fish).—Length without caudal = 3 7/9 that of head (with caudal, 4 ⅓).

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

G. kokopu (smaller fish).—Length without caudal = 3 6/8 that of head (with caudal, 4 ⅓).

G. postvectis.—Length without caudal = 4 that of head (with caudal, 4 ⅗).

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

G. robinsonii.—Length without caudal = 5 2/7 that of head (with caudal, 5 6/7).

G. fasciatus.—Length without caudal = 4 ⅔ that of head (with caudal, 5 ⅓).

– 85 –

G. kokopu (larger fish).—Depth of body under dorsal contained a little more than 4 ½ times in length, without caudal.

G. kokopu (smaller fish).—Depth of body under dorsal contained about 5 times in length, without caudal.

G. postvectis.—Depth of body under dorsal contained 5 ½ times in length, without caudal.

G. robinsonii.—Depth of body under dorsal contained 7 ¾ times in length, without caudal.

G. fasciatus.—Depth of body under dorsal contained 5 ⅕ times in length, without caudal.

G. kokopu (larger fish).—Greatest depth of body contained about 4 times in length, without caudal.

G. kokopu (smaller fish).—Greatest depth of body contained about 4 ⅓ times in length, without caudal.

G. postvectis.—Greatest depth of body contained about 4 ¾ times in length, without caudal.

G. robinsonii.—Greatest depth of body contained about 7 ¾ times in length, without caudal.

G. fasciatus.—Greatest depth of body contained a little less than 5 times in length, without caudal.

In G. kokopu and G. postvectis the length of head is greater than depth of body under origin of dorsal by distance from snout to margin of orbit.

In G. fasciatus the length of head is greater than depth of body under origin of dorsal by half-distance snout to orbit.

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

In G. robinsonii the length of head is greater than depth of body (which is of even depth from back of head to origin of dorsal) by 2/7.

In G. kokopu (small fish), G. postveotis, and G. fasciatus the length of head exceeds the greatest depth of body, but less so in small G. kokopu and fasciatus than in G. postvectis; but in large G. kokopu the length of head equals the greatest depth of body.

G. kokopu.—Diameter of eye = ⅓ length of head, and about 1 ⅓ in snout.

G. postvectis.—Diameter of eye = 5 ½ in length of head, and about 1 ⅓ in snout.

G. robinsonii.—Diameter of eye = little over 8 in length of head, and about 2 ⅓ in snout.

G. fasciatus.—Diameter of eye = 6 times length of head, and about 1 ½ in snout.

G. robinsonii.—Diameter of orbit 7 times in head, and about 2 in snout.

G. kokopu.—Length of pectoral = half-distance of its root from ventral.

– 86 –

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

G. postvectis.—Length of pectoral = 1/16 more than half-distance of its root from ventral.

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

G. robinsonii.—Length of pectoral = 1/14 more than half-distance of its root from ventral.

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

G. fasciatus.—Length of pectoral = 1/12 more than half-distance of its root from ventral.

G. kokopu.—Ventrals terminate ¼ height from vent; extreme length pectorals and ventrals equal.

G. postvectis.—Ventrals terminate ⅓ height from vent; ventrals exceed pectorals in extreme length.

G. robinsonii.—Ventrals terminate more than extreme height from vent; pectorals exceed ventrals in extreme length.

G. fasciatus.—Ventrals terminate about ⅓ height from vent; pectorals exceed ventrals in extreme length.

G. kokopu.—Least depth of tail a little shorter than from end of dorsal to origin of caudal.

G. postvectis.—Least depth of tail ¼ shorter than from end of dorsal to origin of caudal.

G. robinsonii.—Least depth of tail 2 ⅓ times in distance from end of dorsal to origin of caudal.

G. fasciatus.—Least depth of tail ⅕ shorter than distance from end of dorsal to origin of caudal.

G. kokopu.—Anal, if laid backwards, barely reaches base of caudal.

G. postvectis.—Anal, if laid backwards, does not reach base of caudal.

G. robinsonii.—Anal, if laid backwards, nearly twice its height from base of caudal.

G. fasciatus.—Anal, if laid backwards, reaches to base of caudal.

G. kokopu.—Origin of pectorals exactly half-distance to ventrals.

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

G. postvectis.—Origin of pectorals less than half-distance to ventrals by 1/11 height of pectorals.

G. robinsonii.—Origin of pectorals trifle more than ⅖ distance to ventrals.

G. fasciatus.—Origin of pectorals less than half-distance to ventrals by ⅓ height of pectorals, or ⅛ distance snout to ventrals.

G. kokopu.—Origin of pectorals to origin of ventrals much more than half-distance to anal.

G. postvectis.—Origin of pectorals to origin of ventrals more than half-distance to anal.

G. robinsonii.—Origin of pectorals to origin of ventrals very little more than half-distance to anal.

G. fasciatus.—Origin of pectorals to origin of ventrals more than half-distance to anal.

G. kokopu.—Origin of ventrals much beyond midway between length to base of caudal.

– 87 –

G. postvectis.—Origin of ventrals a little beyond midway between length to base of caudal.

G. robinsonii.—Origin of ventrals diameter of orbit less than midway between length to base of caudal.

G. fasciatus.—Origin of ventrals diameter of orbit less than midway between length to base of caudal.

G. kokopu.—Origin of ventrals less than midway total length.

G. postvectis.—Origin of ventrals much less than midway in total length.

G. robinsonii.—Origin of ventrals depth of body less than midway in total length.

G. fasciatus.—Origin of ventrals extreme length base of dorsal less than midway in total length.

G. kokopu.—Distance to origin of pectorals much exceeds distance ventrals to anals.

G. postvectis.—Distance to origin of pectorals a trifle more than distance ventrals to anals.

G. robinsonii.—Distance to origin of pectorals but two-thirds distance ventrals to anals.

G. fasciatus.—Distance to origin of pectorals about same as distance ventrals to anals.

G. kokopu.—Distance to origin of pectorals much exceeds distance origins anal and caudal.

G. postvectis.—Distance to origin of pectorals equals distance origins anal and caudal.

G. robinsonii.—Distance to origin of pectorals four-fifths distance origins anal and caudal.

G. fasciatus.—Distance to origin of pectorals a trifle shorter than distance origins anal and caudal.

G. kokopu.—Extreme length base of dorsal much greater than depth of gape of mouth.

G. postvectis.—Extreme length base of dorsal less than depth of gape of mouth.

G. robinsonii.—Extreme length base of dorsal greater than depth of gape of mouth.

G. fasciatus.—Extreme length base of dorsal equals the depth of gape of mouth.

G. kokopu.—Height anal less than base, but exceeds depth ventrals or distance dorsal to origin caudal.

G. postvectis.—Height anal exceeds base, but less than depth ventrals or distance dorsal to origin caudal.

G. robinsonii.—Height anal exceeds base, but less than depth ventrals or half-distance dorsal to origin caudal.

G. fasciatus.—Height anal equals base, and equals depth ventrals and distance dorsal to origin caudal.

G. kokopu.—Lower jaw projects beyond upper whether closed or open.

– 88 –

G. postvectis.—Lower jaw about same length as upper when shut, but projects beyond when open.

G. robinsonii.—Upper jaw projects beyond lower either closed or open.

G. fasciatus.—Upper jaw scarcely projects beyond lower.

G. kokopu.—The maxillary extends to the vertical from rear margin of pupil of eye.

G. postvectis.—The maxillary extends to about vertical from centre of pupil of eye.

G. robinsonii.—The maxillary extends to the vertical from middle of eye.

G. fasciatus.—The maxillary extends to the vertical from rear margin of orbit.

Galaxias kokopu.

Body very stout; head broad and depressed, slightly flat on top. Fins very fleshy and thick, especially the vertical ones, which also have a thick fleshy pad at their feet, hiding the bases of the rays altogether, first three rays of dorsal and anal being completely hidden. Origin of dorsal slightly in advance of that of anal. Anal higher than dorsal. Ventrals and pectorals each longer than height of dorsal, but less than height of anal. Caudal much rounded. Dorsal immediately over vertical from anal orifice. Width of caudal at base equals three long diameters of orbit or width of head at rear of orbit. Ribs thin, long, and needle-like.

This species has mature ova in November.

Colour.—The ground-tint of sides, back, and head more or less dark pinkish - brown of beautiful transparent tint when the fish is in good condition and alive, but varying in intensity with the class of water frequented. The belly light pinkish-grey, sometimes with quite a golden or bronzed colour. The whole of the sides, back, cheeks, and base of vertical and tail fins more or less closely spotted and marked with rounded spots, and more or less crescent-shaped, and short slightly sinuous or small ring-like marks, some quite old Arabic-script-like in character, and in colour light yellowish-brown or ochre. The spots and marks on the back become much smaller and closer; those on head smaller still, more rounded, and invariably arranged concentrically. The vertical fins and tail are blackish. Pectorals and ventrals lighter. Iris of eye golden, with bluish patches.

Measurements of 11.05 in. specimen are appended.

The Westland people generally call this species the “mountain-trout.”

Galaxias postvectis.

Body is stout, of squarer section than G. kokopu; head also flattened on top, but not so much depressed. Fins are not

– 89 –

very fleshy, the first rays of vertical ones only being buried in membrane. Base of dorsal shorter than that of anal; anal higher than dorsal. Extreme length of dorsal and pectoral equal. Caudal is not so much rounded as in G. kokopu. Vertical from origin of dorsal a little posterior to anal orifice. Bibs are thicker in comparison than those of G. kokopu.

Colour.—Ground-tint pinkish olive-brown, lighter or darker according to class of water. The rearward portion of body only, from about half-way between ventrals and anal, has the sides barred with from 7 to 9 darkish-brown lines, each slightly curved, and thicker along the median line, the apex pointing forwards, the bars diminishing in intensity of colour anteriorly. The rest of the body is unmarked, except a purplish patch bordered with lighter above origin of pectorals. The fins are light-brown, golden-pink in some lights. Iris of eye golden.

This species is not in full roe until December, and is generally called in Westland the “barred trout.”

Measurements of specimen 6.35 in. in length are appended.

Galaxias robinsonii.

Body much elongated and rounded, slightly flattened on back and belly, where the median depressions are well marked. Thickness equalling depth almost to origin of dorsal; from thence to tail the sides are flattened, and body compressed. The eye is small, and set in a much larger orbit. The cheeks are very fleshy and tumid, the opercular divisions being completely indistinguishable thereby without dissection. Caudal fin is slightly emarginate, the angles of lobes, however, being much rounded. The other fins also are very much rounded, and are excessively fleshy and opaque, the general epidermis covering fully two-thirds of their respective free portions. This renders the fin-ray enumeration very difficult without absolute dissection of each ray of the vertical fins especially. The formulae of dorsal and anal externally read 9 and 10 rays, but three first of dorsal and anal are very small and fine, completely imbedded in the dark cuticle, and the last rays of same are very fine and close together. The pectorals are placed very low down on ventral surface, and are quite pedunculated; they originate immediately at termination of gill-openings. Two teeth on each side of lower jaws are much enlarged. The maxillaries have their exposed portions very large and tumid as compared with other Galaxidæ. The anal orifice has a double small papillary appendage. Top of head is wide, but flattened portion is short. Ribs much thicker in proportion than in any other species of this family I am acquainted with.

– 90 –

Pectoral fins are larger than ventrals, and, though the extreme height of dorsal and anal and their basal lengths slightly differ, they are practically of the same size. The origin of dorsal is so far in advance of that of anal that a line arising vertically from last intersects middle of base of dorsal.

Colour.—Ground light bluish-grey, minutely spotted with darker grey and light-brown, and generally spotted and marbled on sides, fins, and tail with dark olive-brown. The spots on back, head, cheeks, and under dorsal (especially the last) are very much the larger. One light patch or semi-band above ventrals and another commencing low down in advance of vertical from origin of dorsal extending back over anal. Belly the general ground colour. Iris of eye golden.

Measurements of specimen described attached.

Galaxias fasciatus.

Last ray of dorsal and anal are double. Longest rays of dorsal exactly equal in height longest rays of pectoral, and equal distance from posterior margin of eye to free margin of gill-covers, equals depth in vertical through posterior margin of eye, and equals the direct distance between the termini of bases of dorsal and anal fins. The longest rays of caudal fin equal the extreme length from superior origin of pectoral to rear free margin of such fin, and also equals the extreme width of body. The eye-opening is inclined to be slightly oval, the larger diameter being the vertical one. The origins of dorsal and anal are in the same vertical. Top of head over eyes slightly rounded, behind eyes flattened. The median depressions along back and belly well marked. The width of gape equals its depth. The commencements of fins are fleshy, the first rays of each being quite imbedded in such, remainder of fins being more generally free from fleshy thickening, and rays approximating in appearance therefore to G. abbreviatus. Cheeks and head are fleshy, hiding opercular divisions. Ribs are much shorter than those of a G. kokopu, postvectis, or robinsonii. In full roe and milt end of October.

Colour.—Rose-brown mottled with darker on the sides and back, and with light-coloured reticulating markings on the back, and 16 or 17 light-coloured irregular transverse large markings or bands along the sides, which are margined on their forward edges with a darker tint. A large irregular bluish marking above the origins of pectorals. Vertical fins darkish and mottled. Iris golden.

Measurements of specimen 8 in. in length appended.

The left-hand branchiostegous curtain overlaps the right-band portion under the throat. Four branchial arches.

Plate IV. shows Galaxias kokopu, 15 ¼ in. and 11 in. total

– 91 –

length; and Plate V. shows G. postvectis, 6.35 in.; G. robinsonii, 8.2 in.; and G. fasciatus, 8 in.—total lengths. The drawings of the fishes are all reduced to the same length, for greater facility in comparisons.

I must apologize for extending my literary matter on the subject to such length, but without doing so it was impossible to bring the various points of difference or resemblance under easy inspection.

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

Galaxias kokopu Galaxias postvectis Galaxias robinsonii. Galaxias fasciatus.
In. In. In. In.
Total length 11.05 6.35 8.2 8.0
Length of head 2.5 1.36 1.4 1.5
Greatest depth 2.1 1.15 1.0 1.40
Greatest width 1.0
Length to base caudal 9.5 5.55 7.5 6.85
Length to origin dorsal 7.25 4.15 5.40 5.15
Length to origin pectorals 2.7 1.36 1.4 1.625
Length to origin ventrals 5.4 2.85 3.6 3.65
Length to origin anal 7.4 4.20 5.7 5.15
Distance pectorals to ventrals 2 86 1.6 2.2 2 05
Distance ventrals to anal 2.15 1.4 2.1 1.5
Depth body at ventrals 2.0 1.1 1.0 1.4
Depth body at dorsal 1.8 1.0 1.0 1.3
Depth body at end anal 1.07 0.7 0.65 0.8
Length base of dorsal 1.0 0.55 0.62 0.70
Length base of anal 1.4 0.7 0.65 0.94
Origin ventrals to base caudal 4.2 2.63 4.0 3.20
Dorsal to base caudal (centre) 1.45 0.85 1.50 1.00
Anal to base caudal (centre) 1.1 0.65 1.30 0.8
Length to centre of eye 0.85 0.46 0.5 0.5
Greatest height of dorsal 1.22 0.8 0.7 0.85
(7th ray) (5th ray) (7th ray)
Greatest height of anal 1.45 0.85 0.8 0.92
(8th ray) (8th ray) (8th ray)
Greatest length of ventral 1.35 0.87 0.86 0.92
(3rd ray) (4th ray) (5th ray)
Greatest length of pectoral 1.35 0.8 1.1 1.06
(4th ray) (5th ray)
Greatest length of caudal 1.50 0.95 0.8 1.14
Width of caudal at base 1.35 0.7 0.95 1.0
Centre eye to pectoral 1.9 0.95 1.0 1.15
Width head at rear of eyes 1.35 0.70
Width head at rear of gill-openings 1.85 0.95
Vertical diameter of eye 0.40 0.25 0.13 0.26
Horizontal diameter of eye 0.45 0.25 0.13 0.25
Diameter of orbit 0.45 0.25 0.22 0.25
Width interorbital space 0.52
Least width of tail 1.07 0.7 0.55 0.8