Lamna nasus (Bonnaterre). Porbeagle shark.
A specimen caught near Otago Heads on May 23, 1933, had three well-developed young within a few weeks of birth, weighing 9 ½, 9 and 7 ½b., a total of 26lb.
Squalus fernandinus Molina. Spined dogfish.
Females are more numerous than males. It appears that while the females are gravid the sexes become segregated, though frequenting the same locality; in March, for example, hauls of males only have been taken in the trawl, and other hauls of females only. The young are born alive during the summer (November to February). Gravid females as small as 24 inches in length have been taken. Small round knobs on the dorsal spines of the embryos protect the parental organs from laceration. In Blueskin Bay, one of the localities most favoured by adults, young specimens evidently born only a short time may be taken in the cod end of the trawl net, so that perhaps parturition,
though usually taking place at greater depths, can occur in these shallow waters. In this bay the young linger for about a year before moving into the Harbour or into deeper water. Possibly birth can occur even in the Harbour, for a female taken at Quarry Point in mid-January, 1930, contained five young in one uterus, all apparently ready for extrusion, with the yolk sac completely absorbed, whereas the other uterus was enlarged and empty as if a half-litter had been born already. The lengths of the five were 10 ¼, 11 ⅛, 11 ¼, 11 ½, 11 ⅝ inches. Another specimen taken in May, 1932, had seven embryos only half as long (4 to 5 ¾ inches) in one uterus, the other uterus being apparently healthy yet sterile.
Raja nasuta Mueller and Henle. Skate.
Capsules containing well-developed eggs apparently almost ready for extrusion were found in October, January and March, suggesting that spawning occurs in the autumn. Two fish two feet in length had developing eggs not yet in capsules. Of numerous capsules taken in the trawl in Blueskin Bay, only one was fertile, and this failed to hatch in an observation tank.
Callorhynchus milii Bory. Elephant-fish. (Plate 41, figs, a, b).
In November, eggs one inch in diameter were found in the ovaries almost ready to pass down the oviduct to be enclosed in capsules; in March, capsules the colour of tow and ready for extrusion were taken from fish caught in Blueskin Bay. On April 19 a capsule was picked up on the beach at Seacliff with an immature embryo 2 ¾ inches long; weight, 10 grams (plate 41, fig. b); yolk sac 1 ¼ inches long and empty. On May 28, a capsule was washed ashore at Brighton containing an almost fully developed fish with distended yolk sac (plate 41, fig. a); weight 320 grains. Empty capsules were often dredged and trawled in Blueskin Bay, no doubt one of the breeding grounds of elephant-fish. The egg cases washed ashore are rarely more than 10 inches long, while those taken from fish in January and from the cod-end of the trawl in February measured 16 ½ inches; colour yellowish-green and almost covered with tow-like fibre. The pointed ends of trawled capsules were discoloured and macerated, suggesting that they had been thrust into the soft ocean floor by the female.
Sardinia neopilcharda Steindacher. Pilchard.
Well developed ovaries were found in numerous specimens from Beaconsfield Bay on December 15, 1931, perhaps exceptionally late. Lengths of fish were as follows: October, 2 ⅛ inches (Cape Saunders, stomach of red cod); February, 3 inches (Beaconsfield Bay); March, 2 ¼ inches (Beaconsfield Bay; evidently a later spawning); March and April, 2 ¼ -2 ¾ inches (Lower Portobello; sexually immature); autumn mature open-sea fish 4–5 inches.
Clupea antipodum (Hector). Sprat.
One ripe female taken from stomach of whiptail caught at The Rock in 50 fathoms off Cape Saunders, May 16, 1931; weight of ovaries, 17 grams; diameter of egg, 0.05 mm. Numerous specimens from the same fish and also bream, same locality, showed the process of spawning had been recently completed. It seems evident that The Rock is in the vicinity of a spawning ground of this fish.
Hippocampus abdominalis Lesson. Seahorse.
A female and three males were kept in a tank during 1931. Ordinarily during the day they remained attached to the outlet pipe or weed, but on December 12 they were unusually active, the males passing and repassing the female as closely as if trying to touch her or approach as near as their feeble movements would allow. When the female, who was evidently ready to transfer her eggs, was temporarily removed from the aquarium, a number of eggs were secured by gentle pressure on the pouch. They were oval, 2.3 mm. by 1.4 mm., semi-transparent, brick red, granulated, and had a distinct perivitelline space. They were sufficiently plastic to conform in shape to the surface on which they rested. They sank in water and had no oil globule.
The female showed marked preference for one of the three males, which swam almost erect, jerking their heads forwards as they approached. When the two unfavoured ones approached, the female turned away as far as possible, or if they were too attentive, released her hold on the weed and swam off, if possible to her favourite. The removal of the latter from the tank did not restore the others to favour; he was therefore replaced, and the other two removed. Even then, after many approaches by the male and after various stationary and swimming postures by the female, there was considerable difficulty in securing the requisite nuptial embrace. All the movements were slow, though somewhat less deliberate than normal in this sluggish fish. The required grasp was at length secured while the female had her tail resting on the floor of the tank, and in an instant the oviduct, which appeared enlarged, was inserted into or at least against the pouch of the male and some of the eggs were transferred. This inverted coitus lasted a few seconds, after which the pair separated, the female making violent efforts to free herself, and, when successful, attaching herself to a stalk of weed; the male performed contortive movements on the floor of the tank, stretching and twisting as if to arrange the eggs in place. At intervals of five minutes, at least 13 embraces occurred; in the intervals the flaps of the male pouch were observed to be dilatable and apparently under muscular control. After the last embrace, when the male pouch had become distended and the female compressed, the male attached himself to the outlet pipe of the tank and remained there apparently without feeding for many hours, after which he began to feed and relished a supply of amphipods (Caprellina longicollis) placed in the tank. The oviduct of the female receded considerably during the next few hours; the outer part is an intromittant organ under at least temporary control. It was not determined whether fertilisation occurs during the process of egg-transference.
Males are ovigerous in summer, as verified several times by dissection. The eggs hatch in January and February, 30 days after transference, the young being fully developed and able to swim, averaging 17 mm. in length; up to 182 are “born” from a single parent. The latter process was observed several times; it occupied about an hour, the young emerging rapidly in succession with occasional interruptions, during which the male rubbed the pouch
on the weed as if to press the remaining ones forward. Prolonged observations did not support the assertion that in this genus the young are able to enter and leave the brood sac at will. For the first few hours the young swam parallel and close to the surface of the water, and were attracted by light, but later sank to any depth.
Hemirhamphus intermedius Cantor. Garfish. (Plate 42, figs. 1–11).
Specimens nine inches in length and weighing only four ounces were found to have ripe roes.
The first attempt to hatch artificially-fertilised eggs, from parents; taken with he seine net in Company's Bay, November 23, 1931, failed after a fortnight, through the deposition of rust from the iron water-supply pipes. A month later, another lot progressed satisfactorily after the substitution of rubber hose and the exercise of meticulous, care in removing sediment and dead eggs. Hatching boxes with an adequate circulation of water proved more successful than glass jars. The relatively large eggs, diameter 2.6 mm. (plate 42, fig. 1), are demersal, and in the tanks became attached to weed by the filaments, or collected into balls of several dozen, which had to be shaken apart to free then from sediment. Eggs were noticed on weed in the stomachs of mullet caught in the same haul as the parent garfish.
Stages in the development are shown in pl. 42, figs. 1–11; the period of incubation ranges from 36 to 45 days. In the one-day fish, the dorsal fin commences above the vent, which is at the posterior two-thirds of the length; the dorsal, caudal, anal and ventral fins are continuous. The characteristic silvery colour of the adult has been assumed, but the caudal fin is not yet forked and the lower jaw not yet elongated (figs. 9–11). This fish (fig. 9) shows the lower jaw-less than 0.1 mm. in length; in a two-day-old fish (fig. 10) the jaw measured 0.15 mm., and after four days (fig. 11) 0.25 mm. The only larva taken, after repeated attempts, in a tow net measured 26 mm., the lower jaw 4 mm. in length.
Coelorhynchus australis (Richardson). Javelin-fish.
One specimen, ovaries almost fully developed, August 26, 1931; eggs 37,000, diameter 0.35 mm.
Macruronus novae-zelandiae (Hector). Whiptail.
One specimen, ovaries almost ripe in May.
Merluccius gayi (Guichenot). Hake.
A specimen taken at North Reef in 200 fathoms (July 25, 1933) contained about 1,450,000 eggs; weight of roes 275 grams; eggs bright orange, diameter 0.06 mm.
Physiculus bachus (Bloch and Schneider). Red cod.
The roes are ripe during July and August, at which times the fish do not bite freely. The sexually mature adults leave the harbour and other inshore grounds at this time, and for a short period may be taken in 20–23 fathoms. The presumption that they spawn in deeper water, beyond the range of the trawlers, is supported by the catching of ten fish with ripe roes at the North Reef in 80 fathoms. (September 16, 1931); these fish, which were in good condition, averaged 29 inches in length and 7 ½lb. in weight; the roes varied from 7 ½ to 8 ½ fluid ounces, and as shown in the following table, the eggs varied considerably in number and in size.
|Number of eggs.||Average diam. of eggs.||Diameter of oil globules.|
|4,900,000||0.35 mm||0 08 mm.||0 02 mm.|
|3,350,000||0.40 mm.||0.10 mm.||0.05 mm.|
|2,600,000||0.45 mm.||0.13 mm.||0.07 mm.|
Assuming that red cod spawn in September, the juveniles caught in Blueskin Bay and the Harbour in March, from four to six inches long, are about six months old, and those taken in August and September in the same localities, measuring about nine inches, are one year old. Specimens 12 to 15 inches long, common off the northern side of Goat Island in the Harbour in late summer and early autumn, are presumably two years old. Sexually mature fish measure at least 20 inches and are probably four years old.
Pseudophycis breviusculus Richardson. Bastard red cod.
A ripe female was taken among blue cod at Taieri Mouth (January 10, 1933), at least four months later than the end of the spawning season of red cod. The eggs were estimated at 2,260,000; diameter 0.5 mm., oil globule 0.15 mm.; eggs larger than red cod, globules larger but less numerous. Recently spent fish were secured in February, 1933.
Zeus faber Linnaeus. John Dory.
One specimen with mature ovaries taken in August, 1933, eggs about 37,000; diameter 1.8 mm., white, almost transparent; oil globule conspicuous, lemon coloured, diameter 0.32 mm. Specimen in poor condition with numerous crustacean parasites on the gills.
Cyttus australis Richardson. Boarfish.
Ovaries not fully mature in September, probably ripe about October.
Cyttus novae-zelandiae (Arthur). Silver dory.
Eggs well developed in August; diameter 0.75 to 1.0 mm.
Peltorhamphus novae-zelandiae Guenther. Common sole.
Maturation is evidently a long process; the ovaries begin to enlarge in the small fish, and immature eggs were found in a 6 ¼ inch specimen; on the other hand, the smallest specimen with mature eggs measured 9 ½ inches. No such sign of maturity was found in specimens from the Harbour, yet juveniles ranging from 1 ½ to 6 inches are common there throughout the year.
Agonostomus forsteri (Cuvier and Val.). Yellow-eyed mullet.
The ovaries are well developed in December. Juveniles 1 ¼ inches long are common close inshore in the Harbour, where they are preyed on by 9-inch barracouta.
Hyperoglyphe porosa (Richardson). Bream.
Roes commence to enlarge in September and are almost ripe in December.
Polyprion americanus (Bloch and Schn.). Bass-groper.
The roes are ripe in July; eggs 11,600,000, diameter 0.22 mm., oil globule 0.05 mm.
Polyprion oxygeneios (Bloch and Schn.). Groper.
Although the early stages remain unknown, no specimens less than one pound in weight having been seen, the movements of the parents are known. In May and June, when the roes mature there is an outward movement into deeper water such as the North Reef to spawn. The migration is well defined, and is recognised by fishermen, who follow the fish as they retreat from the coast. Spawning occurs in July and August, at which season the females preponderate in the catches in the ratio of 3:1. A discovery of ripe roes late in December was evidently quite exceptional. In October and November the return movement commences, the fish, especially the females, being in their lowest condition. During July and August the bottom jaws, the lips, the nose and the undersurface of the shoulders of mature fish are scratched or scarred. It may be that this results from a habit of nosing or scratching away sand, gravel or rocks in order to spawn (Tenison-Wood, 1883, p. 103) or from fighting with one another at spawning time.
The prevailing wanton methods of fishing, including over-fishing, have intensified exploitation during the spawning season, resulting in a serious is decline in the numbers and sizes of the fish, and action is called for with a view to gaining control over the situation. At best, recovery from the damage already done will probably be slow. An indication of the destructive effect of the present lack of control was obtained in 1932 and 1933 by estimating the number of ripe female groper marketed from the North Reef. Many of the ripe roes weighed 9lb., some even 10 ½lb. Roes containing a million eggs per pound weight were seen, and even half this number would imply that 300,000 million eggs were sold on the market during each year. Such a lost, appalling in its magnitude, even in a marine environment, cannot be viewed with equanimity.
Trachur is novae-zelandiae Richardson. Horse-mackerel.
Ripe males and females were examined on the market (February 22, 1933). Length of female 20 inches, weight 2 ¾lb.; diameter of eggs 0.6 mm., yolk 0.4 mm., no oil globule; total eggs 440,000.
Arripis trutta (Forster). Kahawai.
Roes veil developed in February, probably maturing in March or April.
Dactylopagrus macropterus (Forster). Tarakihi.
Specimens with ripe roes were taken at the North Reef in 100 fathoms (December 8, 1932); the ovaries of a 23-inch fish (cleaned weight 7lb) weighed slightly over four ounces and contained about 3,300,000 eggs; colour orange, diameter 0.2 mm.
Latris lineata (Forster). Trumpeter.
The roes are ripe in July, and the fish are in optimum condition in May and June, poor in July after spawning. Ripes roes from 25lb. fish in July 1933, weighed 4 ounces and contained about 12,000,000 eggs. The winter spawning occurs in water well below the 23-fathom line to which the fish migrate in summer.
Coridodax pullus (Forster). Greenbone.
Repeated attempts at artificial fertilisation, between November 9, 1931, and January, 1932, at the station failed because no ripe males could be secured. A 19-inch female had 67,000 eggs; diameter 0.23 mm., granular and pink in colour. There is no ready explanation of this apparent protogyny, unless the males move into deeper water when fully mature, the females following.
Crapatalus novae-zelandiae Guenther.
A specimen examined in June was almost ripe with 18,000 eggs; diameter 0.47 mm.
Parapercis colias (Forster). Blue cod.
The departure of gravid fish in June and the return of spent fish in September indicate the normal limits of the spawning season. In 1933 gravid females were nevertheless found in mid-December, perhaps after an ineffective visit to the spawning grounds. The roes of a 2 ½lb. specimen of these late-maturing fish contained about a million eggs; diameter 1.01 mm.
Katheostoma giganteum Haast. Flathead.
Ova well developed in February, almost mature in March. Those of a specimen weighing 10 ¼lb. contributed 2lb. to this weight, and contained some 800,000 eggs.
Notothenia macrocephala Guenther. Maori Chief.
Eggs ripe in November; diameter 0.25 mm.
Thyrsites atun (Euphrasen). Barracouta.
The late autumn period of April-May-June, during which the fish reach their best condition, is the period of the development of the roes; towards the end of July they are nearly mature and the fish disappear from the surface. The interpretation of this disappearance is a matter for constant debate among fishermen; it is not known whether there is a minor downward movement, a movement to the inshore bottom, or an offshore migration by the main body at least. Some certainly remain in comparatively shallow water, for mature specimens are occasionally taken in winter on the line, e.g. from The Rock at Cape Saunders and in 80–100 fathoms on groper beds 12 miles offshore. The fact that only a few are taken at these depths after the July migration does not disprove the presence there of large numbers, especially as those taken are never found to have been feeding freely; in October and November the stomachs are completely empty, or at most contain a few benthic fishes and crabs; it is a perplexing detail that these include an admixture of inshore benthic animals.
Until it is more reliably known whether the main body is in the neighbourhood of this 100-fathom zone in winter, it cannot be claimed that the spawning grounds are known. However, there can be no doubt that the disappearance in July is a spawning migration, as sufficiently shown by the prior maturation of the gonads, and by the reappearance of the fish in numbers at the surface in October, when they are found to have spawned and to have become very thin and hungry.
The data as to the movements of young fish are still more fragmentary, especially since the conspicuous inward migrations in summer and autumn, when extensive shoals enter the Harbour, are performed by juveniles of three sizes (4, 9 and 18 inches), so sharply marked off from one another as to indicate three ages at distances of a season. These have evidently come from the same bottoms as the adults that are caught during the spawning season, for on the supposed spawning grounds young specimens 4–7 inches long are found in the stomachs of several species of fish, including red and blue cod, kingfish and sea perch.
Acanthoclinus quadridactylus (Bloch and Schn.). Rockfish. (Plate 41, fig. c; plate 42, fig. 12.)
Eggs were found at Seal Point, Otago, in October, 1931, but the following observations were made at Auckland in 1926. In July and August the ova were observed to be nearly mature; in October a female was found under a stone in an intertidal cavity containing Zostera leaves, and was evidently in the act of spawning. The tail was quietly moving, and the fish appeared unperturbed by the temporary removal of the stone. A fortnight later the cavity was found to be occupied by a male, guarding a ball of eggs; the sides had been built up so that neither fish nor eggs could escape, nor could enemies enter. The water was not sufficient to cover more than half the bulk of the eggs or the fish. From October to December several other such cases of self-imprisoned males in charge of a ball of eggs were found. In marked contrast to their usual behaviour the fish were unperturbed by the removal of the eggs, or even by the handling of the fish itself. The tail quietly lashed the water, ceasing when the eggs were lifted and recommencing when they were replaced. Eggs were found in different stages, and in one case they had already hatched, the larvae swimming in the pool still accompanied by an adult male. It appears that, one or the other sex, probably the female, chooses a suitable spot and scoops out a shallow hole without closing the entrance; the female then deposits the eggs and departs; the male then enters, closes the entrance, fertilises the eggs, and remains on guard until after they are hatched. The heavy egg cluster (plate 41) was held together by threads, which also entangled Zostera leaves, sponge and other extraneous matter, and formed a mass sufficiently compact to be picked up and suspended from a piece of wire. Shortly before hatching (pl. 42, fig. 12), the embryo is large and plump, occupying the whole circumference of the egg, and with the circulation well established. The posterior portion of the body is free from yolk and the embryo moves freely within the egg. The pigmentation characteristic of the adult is already in evidence. Diameter of egg 1–3 mm., oil globule 0.12 mm.
A fish was noticed to spend much time in and around a gastropod shell (Lunella smaragdus), and on the following day appeared even more attentive, spending up to 20 minutes at a time in the shell; after a period inside, with tail or head or neither visible in the aperture, it would partly emerge and perform regular movements with the fins and tail. It was later verified that eggs had been deposited within.
A fortnight later the female was seen chasing a smaller fish, evidently a male, along the walls of the pond and among weed, until they began to encircle the shell in narrowing circles and finally came to rest on it. After about a minute the male made a brief inspection of the interior, returned and rested on the shell with the female, then entered the shell again and did not return. The female thereupon withdrew and, after feeding ravenously on pieces of fish thrown into the water, departed and was not seen again. The male stayed in the shell, sometimes with the head protruding, and was not observed to leave the shell or to feed. After another fortnight the shell was raised with a net, the male remaining placidly within, and observations were continued in a smaller aquarium. The fish was seen to dart into the shell (plate 43, fig. a), turn round, and partially emerge with the head protruding (plate 43, fig. b); then violent movements from side to side were seen, with rapid movements of the pectoral and ventral fins; then the fish emerged further, and the tail could be seen lashing the water inside the shell. Diameter of egg 0.9 to 1 mm.
Genypterus blacodes (Bloch and Schn.). Ling.
Well-developed roes occur in August with very numerous eggs; ripe in late September, weight 116 grams, eggs approximately 1,200,000; diameter 2.1 mm., oil globule 0.4 mm.
The vivipary already recorded by Anderton (1921, p. 88) adds especial interest to the development and breeding of this species. The two ovaries (pl. 45), varying in size from 70 × 38 mm. to 60 × 25 mm., are contiguous with the duets coalescing anteriorly; the progressive formation of abundant oil globules tinges the originally colourless ovary until it reaches a yellowish-grey hue, overlain ventrally by dark grey or blackish pigments in the peritoneum. The highly vascular ovarian membrane is characterised by transverse ridges. The ovarian follicles (pl. 46, fig. 25) are stalked, terminal or secondary or tertiary or further subdivisions of a complex branching system, richly vascular, and provided with a granulated periphery inside the smooth enveloping capsule. Some eggs lie free in the semi-fluid ovarian matrix, including some which are larger than any of the stalked ones; but detachment does not depend on the attainment of maturity, as shown by the great range of stages among the stalked eggs.
This range is such as to suggest that parturition is intermittent; it can at least occur in two phases, for in a single ovary there may be found not only abundant examples of all the stages represented in a continuous series (pl. 46, figs. 26–34), but also, after a marked gap without further intermediate stages, the full-term larvae ready for extrusion (fig. 35). In another case one ovary contained 80,000 eggs ranging in size from 0.25 to 0.75 mm., and the other had a similar number of larvae ready for extrusion. The finding of this latter case in February, 1930, and of another similar one in June, seemingly indicates an extension by six months of the spawning season as recorded by Anderton (1921, p. 88), who found such specimens in September and December. Unless some of these cases are rare anomalies, there can be at most no more than a brief non-spawning season in July and August.
The process of spawning was witnessed in February, 1930. After a restless period contrasting with the usual sluggishness, a captive seaperch turned partly on one side and for about half an hour extruded at ten-minute intervals jelly-like material which soon floated to the surface; the total mass occupied 155 cc. and contained some 88,000 larvae (pl. 46, figs. 35, 36, 37), wriggling in the jelly-like matrix, which dissolved in about 20 minutes. Next day (pl. 46, fig. 39), the young fish were swimming freely, mostly at an angle of 45° in the water; the jaws were as yet imperfectly formed, and it was not until the fourth day (pl. 46, fig. 40), after the disappearance of the yolk sac, that plankton was observed to be eaten.
Congiopodus leucopoecilus (Richardson). Pigfish.
Roes were observed in June, there being over 50,000 eggs per fish. In spite of the constant use of the garfish bunt during seine netting operations in the Harbour, no pigfish smaller than 4 ½ inches were taker. In captivity a fish of that size increased one inch in length in live months (November to April)), probably slower than in its natural habitat.
Neophrynichthys latus Hutton. Toadfish. (Plate 47, fig. a.)
Spawning occurs in July and August; eggs, 13,000; diameter 2.6 mm. In the observation tanks eggs were deposited on a piece of rough rock, to which they adhered (see figure), or were scattered about in small heaps of several hundreds (they remained unfertilised, as no males were available).
Chelidonichthys kumu (Lesson and Garnot). Red gurnard.
Gurnard usually spawn in Search. A 14 ½-inch fish (marketed December 6, 1933) had fully mature roes with 180,000 eggs; diameter 1.7 mm.; copper-coloured oil globules, 1.3 mm.; sometimes coalescent after spawning. About 80 per cent. of some hundreds of fish examined were females. The smallest ripe female measured 11 ½ inches.
Diplocrepis puniceus (Richardson). Sucker-fish. (Plate 46, figs. 41–45; plate 47, fig. b.)
The secondary sexual differences are striking. Apart from the much smaller size, males are distinguished by the brilliant colours; the females are bright rose on the upper surface and yellowish below, whereas the males are olive-green, brown or grey with darker lines, tinged with red, yellowish or pale lilac above and black laterally; others are red with pale and dark longitudinal stripes. There is an endless variety in the forms and arrangement of the colour patches, some males being striped, others being spotted, others blotched and barred. The colours of the males are even brighter in the breeding season, when they are more frequently seen. In other seasons they appear outnumbered by the females in the ratio of 5:1, but on account of their small size and presumed preference for deeper water it is doubtful whether there is any genuine disparity in the relative numbers.
Eggs were seen from early September to the end of November, deposited in irregular masses under rocks (plate 47), stones and shells, almost invariably on smooth surfaces. They are usually found a little above low-water mark, sometimes higher up, in pools which
Fig. 1.—Unfertilised egg, diameter 2.6 mm., adhesive filaments 1.7 mm. Oil globules numerous, central (us shown) or scattered. Fig. 2.—Egg 26 hours after fertilisation; blastoderm 5-celled. Fig. 3.—Egg 23 days after fertilisation; embryo well formed, with pigment spots; less than periphery of egg; no movement or circulation observed (unusual). Fig. 4.—Egg 26 days after fertilisation; embryo forming complete loop, moving freely within membrane; pigmentation conspicuous posteriorly, but embryo relatively transparent. Fig. 5.—After 30 days; pigmentation further advanced; yolk sac and oil globule still recognisable. Fig. 6.—Embryo ready to hatch. Fig. 7.—Newly-hatched larva; length 7 mm., yolk sac still large, oil globule still present; head not deflected, dorsally pigmented. Fig. 8.—One-day-old larva; length 8 mm.; tail fin with rays raised, myotomes distinct, oil globule and yolk sac reduced; pigmentation especially prominent above head. Figs. 9–11.—Dorsal views of head, ages 1, 2, and 4 days respectively, showing commencement of elongation of lower jaw (lengths 0.1 mm. or less, 0.1 mm. and 0.25 mm. respectively). Fig. 12.—Un-hatched embryo taken from egg cluster, pl. 41, fig. c—Acanthoclinus quadridactylus (Bloch and Schn).
Fig. 13.—Ripe egg. 0.9 to 1 mm.; scattered oil globules and yolk; internal ventral filaments. Fig. 14.—Segmenting egg; embryonic shield surrounding oil globules; embryonic axis defined. Fig. 15.—Ventral attachment filaments. Fig. 16.—Central oil globules; micropyle. Fig. 17.—Embryo exceeding a complete circle; pigmentation slight. Fig. 18.—Do., showing dorsal pigmentation. Fig. 19.—Do., with adhesive filaments. Fig. 20.—Do., chromatophores and oil globules. Fig. 21.—Embryo ready to hatch; pigmentation well marked, eyes green, yolk sac slightly reduced; the larva turns over about 20 times per hour. Fig. 22.—Newly hatched larva; length 3.8 mm., yolk sac incompletely absorbed and hindering movements; oil globules 7, diameters 0.05-0.17 mm.; green dots behind yolk sac, red jagged patch above eyes, black stellate chromatophores near vent. numerous dorsal and ventral dots from vent to tail. Fig. 23.—Four-day larva; length 5.9 mm.; 2 small oil globules, yolk sac nearly absorbed. Fig. 24.—Six-day larva; length 6.5 mm.; mouth, branchiostegals, and pectorals well developed; head blunt (ef. adult); and bladder conspicuous. Larva able to swim and feed.
Fig. 25.—Follices on secondary and tertiary axes. Fig. 26.—Homogeneous doubly-contoured egg. Fig. 27.—Opaque coagulation in formalin. Fig. 28.—Unsegmented egg, showing micropyle. perivitelline space, scattered protoplasmic granulations. Fig. 29.—Blastoderm cleavage, periblastic nuclei numerous and conspicuous in life. Fig. 30.—Earliar stage, incipient blastodisc in lateral view, showing polar concentration of protoplasm; perivitelline space distinct. Fig. 31.—First on globule at late stage of blastodisc formation. Fig. 32.—Embryonic rudiments at pole opposite the blastodisc. Fig. 33.—Early embryo, occupying ¾ of circumference. Fig. 34.—Minute chromatophores towards tail (typical but not invariable), eyes densely pigmented. Fig. 35.—Larva at time of hatching: uncoiled length 2.2 mm. Fig. 36.—Newly-hatched larva in oblique, dorsal view; body plump, yolk sac protuberant, jaws unformed, posterior dorsal and ventral edges normally with marked pigmentation. Fig. 37.—Larva half an hour after birth, uncoiling, yolk sac reduced; groove in position of future jaws. Fig. 38.—Four-hour larva, feebly swimming, length 2.8 mm. Fig. 39.—One-day larva; length 2.9 mm.; yolk sac almost absorbed; jaws, branchiostegals, and vent recognizable; yellow pigmentation from vent to eye; a patch of black stellate Chromatophores above vent, another towards the tail; two small oil globules in front of original globule. Fig. 40.—Four-day larva, length 3mm., pre-anal granulations especially concentrated on mid-dorsal line; black chromatophores massive posteriorly; myotomes well formed; jaws functional, larva feeding on plankton.
Diplocrepis puniceus (Richardson).
Fig. 41.—Newly-deposited egg; diameter 1.6-1.7 mm.; oil globule 0.42-0 47 mm. Figs. 42–43.—Do., more advanced. Fig. 44.—Ten-day embryo: active, almost encircling periphery; body yellowish; tip of snout and back and especially mid-dorsal line with black stellate chromatophores; oil globule 0.5 mm. Fig. 45.—Newly-hatched larva; length 5 mm., oil globule 0.15 mm.; body compressed, everywhere closely covered with black and brown chromatophoras; yolk sac almost absorbed; head fairly deep, ventral outline angulated mouth a little superior; abdomen well defined with tapering pointed tail.
are not drained at low tide; perhaps also they occur below low-tide level. In the tanks, spawning occurred twice in empty Haliotis shells and once on glass. Much movement occurred during oviposition; after a few minutes of quietness the fish would extrude an egg, then wriggle from side to side as if attaching it.
Parental care was observed on the part of the male. A female interrupted during spawning would endeavour to escape, and in any case the eggs were forsaken when laid; but the male, after having fertilised and assumed charge of the eggs, guarded them assiduously with little concern for his own safety. Male guardians studied in the tanks fasted and became thin and exhausted. They moved constantly back and forth across the mass of eggs, perhaps to keep them free from foreign matter by the fanning action of the tail. In the two cases studied the male was ultimately eaten by a female, presumably his mate. It is uncertain whether this cannibalism was a perversion induced by the captive condition; but, in any case, to swallow, head first, a wide-bodied fish, half as long as the consumer, proved a difficult task. The male was alive, in one case, an hour after capture, the head being out of sight, but the tail protruding from the mouth of the female. After ultimately swallowing the male, the female was much distended and next morning was found dead. Captivity, whether on account of confined space, stillness of the water or high salinity, certainly interferes with oviposition. Fish distended with eggs or even captured in the act of spawning were unable, with the exception of some from sheltered localities, to spawn in the tanks, even though they appeared to be trying to do so. They nevertheless became less distended, and dissection showed that the ova were disintegrating; on the other hand, none of the many sucker-fish kept in the tanks over long periods died unless by violence. The 3 ½-inch fish produced 16,000 eggs, at first soft, but soon hardening. In the tanks about 1 to 2 per cent. were not fertilised. As development proceeded the pink colour was replaced by straw, pale green, and finally purple with progressive intensity. One lot of eggs simultaneously exhibited the whole sequence of colour phases, suggesting progressive fertilisation. Normally the larva hatches after 21 days. Diameter of egg 1.6 to 1.7 mm., oil globule 0.42 to 0.47 mm.