IV.

CONTRIBUTIONS FROM THE NEWPORT MARINE LABORATORY, COMMUNICATED BY ALEXANDER AGASSIZ.

XIV. ON THE DEVELOPMENT OF SOME PELAGIC FISH EGGS. - PRELIMINARY NOTICE.

BY ALEXANDER AGASSIZ AND C. O. WHITMAN.

Read May 14th, 1884.

CONSIDERABLE attention has in late years been paid to fish eggs found floating on the surface, more especially since the establishment of various Fish Commissions to study the history of the sea-fishes. As early as 1868, Malm1 raised the eggs of a species of Flounder by artificial fecundation, and found them to float on the surface at first, and in the later stages of development to sink gradually below the surface. Sars,2 in 1869, found that the eggs of the Cod floated on the surface. Haeckel, during a visit to Corsica, in 1874, also found pelagic fish eggs which he referred to some Gadoid: he subsequently found the same eggs at Nice, in 1876. E. van Beneden, in 1874, studied pelagic fish eggs at Villa Franca. Kupffer,5 in 1868, published some interesting investigations on pelagic fish eggs found in the harbor of Kiel.

Mr. Ryder and Colonel McDonald of the United States Fish Commission observed that the eggs of the Spanish Mackerel were found floating on the surface.

In 1879 and in 1882 A. Agassiz published some preliminary results on the pelagic fish eggs he had raised during the past twenty

1 Malm, A. W. Svenska Vetensk. Akad. Handl., VII., 1867 and 1868.

2 Sars, G. O. Indberetninger til Departementel for det Indre. Christiania, 1869.

8 Haeckel, E. Die Gastrula und die Eifurchung. Jena Zeitschr., IX., 1875.

4 Van Beneden, E. Quart. Journ. Mic. Sci., 1878, p. 41.

5 Kupffer, C. Archiv für Mikr. Anat., 1868, p. 209.

6 Ryder, J. A.

7 Agassiz, A.

Bull. U. S. Fish Com., I., p. 136, 1881.

Proc. Am. Acad, vol. xiv., 1878, p. 1, and vol. xvii., 1882, p. 27.

years, some of them belonging to the Flounder, to Ctenolabrus, to Cottus, to Lophius, and to Tautoga.

9

8

According to Emery, the eggs of Fierasfer are also pelagic. Kingsley and Conn have also studied the pelagic eggs of Ctenolabrus; they state that Mr. Van Vleck has observed those of Merlucius, and also figured an egg with an oil globule; and finally Hensen 10 has published a most interesting paper on the occurrence of the eggs of a few of the fishes of the Baltic at the surface.

During the season of 1883 a good deal of the difficulty, and consequent confusion, existing in distinguishing the many species of pelagic eggs met with during the summer at Newport, has been evercome, and we are now able to distinguish no less than twenty-two species of pelagic eggs, nearly all of which have been referred to some of the many young stages of osseous fishes which have been collected at the surface for a series of years.

The differences between these pelagic eggs are very slight, and the greatest possible care is necessary not to confuse eggs of very dissimilar fishes. I may give as an instance that the eggs of Ps. melanogaster and of an undetermined Flounder had, till this year, been confounded with those of Ctenolabrus; those of the Brown Flounder with those of two species of osseous fishes as yet undetermined; those of a species of undetermined Flounder with those of Hemitripterus; and those of the Sienna Flounder with those of the Yellow Flounder. As these different pelagic eggs must go through extensive changes in the course of their development, changes in the appearance and growth of pigment spots of the body and of the yolk, it is almost impracticable, without any extensive series of sketches, to establish with certainty the identity or difference of closely allied eggs. Hensen has called attention to the ellipsoidal shape of some of the pelagic fish eggs; this is particularly striking in the egg of an Osmerus (?); in this the difference between the longer and the shorter axis can be detected by the eye. The yolk mass of this egg is remarkable for being segmented in large polygonal cells; a similar, but incomplete segmentation occurs in the eggs of the Brown Flounder.

The pigment spots of the surface of the yolk, and those characteristic of different species of fish embryos, begin to make their appear

8 Emery, C. Fierasfer. Arbeit aus d. Zool. Station, zu Neapel.

9 Kingsley, J. S., and H. W. Conn. Mem. Boston Soc. Nat. Hist., III., No. VI., 1883.

10 Hensen, V. Bericht der Com. zur Wiss. Untersuchung der deutschen Meere, IV. Kiel, 1883.

ance at very different times in the many species we have examined. Hence, until the characteristic pigment pattern of an embryo is pretty well known, it is easy to confound the eggs of very different species. The position and shape of the otoliths and the degree of development of the pectorals become also excellent guides to the identification of eggs well advanced in their development. The differences in the young embryos on hatching are very considerable, and in these earlier stages the degree of development of the head, the proportional size of the yolk-bag, the shape of the embryonic fin, the position of the vent, and the pattern of the pigment spots, are all of great use in the identification of the species.

It is remarkable that no monstrosities have ever been picked up among the large number of pelagic eggs examined during the past twenty years, while among the eggs raised by artificial fecundation, the number of eggs which do not develop is very considerable. It is true, that unfertilized eggs, after a day or two, probably fall to the bottom, and are rapidly decomposed, or eaten by other animals. As the majority of the species of Flounders, of which the pelagic eggs have been found, live together in considerable numbers, it is probable that at the time of fecundation but few eggs escape being fertilized: the same is the case with Ctenolabrus, Tautoga, and other shallowwater species. The pelagic eggs are, of course, at the mercy of the winds and waves, and are found in the greatest abundance in the streaks formed by tidal eddies and by winds, which are everywhere on the sea-coast such excellent collecting ground for embryos of invertebrates and for other pelagic animals.

We have collected at Newport the pelagic eggs of six species of Flounders, two species of Cottus, those of Ctenolabrus, Tautoga, Osmerus, and Lophius, and have in collection the eggs of ten species of fishes as yet not determined; but they are probably the eggs of Motella, of Labrax, of Poronotus, and of the Bluefish. The exact identification of these eggs must be deferred to another season. Several of the eggs have been referred to the species of Flounder and other young fishes which were figured in former papers of Mr. Agassiz in the Proceedings of the American Academy of Arts and Sciences.

The presence or absence of an oil globule is an excellent guide in the identification of the egg; the size of this globule is, however, quite variable. In one of the species of Cottus there are many globules present, and the number of these varies from sixteen to thirty-two for this species. In another species, Hemitripterus, in which there is generally only one globule, it is not an uncommon occurrence to find two globules.

Closely allied species of Flounders are found to have eggs either with or without an oil globule. The question naturally arises how far in one and the same egg the number of globules may vary. I have followed an egg in which in some stages the number of globules varied in number from day to day. These pelagic eggs all appear to have a great number of minute fatty globules scattered through the yolk mass; these may or may not unite in a single or in many globules, or may always remain scattered in the yolk. It is undoubtedly to the presence of these minute fatty globules and the larger oil globules that the pelagic eggs owe their capacity for floating. Many pelagic eggs undoubtedly sink in the latest stages of growth.

The number of these pelagic eggs is very great. Hardly a day passes when the fishing with the surface-net does not bring in a number of eggs. The spawning season of many of the fishes which lay pelagic eggs is not very long: at any rate, the different eggs succeed one another quite rapidly, and of the twenty-one species of pelagic eggs thus far observed at Newport, none extend over a greater period than six weeks. The statement I had made, that the eggs of Ctenolabrus were found during the whole summer, rests on the incorrectness of the identification of pelagic eggs closely resembling those of Ctenolabrus, and which are collected in the last part of July and during August.

As the data for the exact determination of some of these eggs are still incomplete, we defer publishing them until they can be supplemented with the observations of another season.

During the summer of 1883, our attention was directed mainly to the earlier stages of development, embraced between the fecundation of the egg and the complete formation of the embryo. The numerous researches on the embryology of the teleostean fishes leave many points of fundamental importance yet to be settled. In evidence of this, we need only refer to the parablast theories that have appeared since the investigations of His; the contradictory views concerning the origin of the so-called "free nuclei" which appear beneath the blastoderm, and the part they play in building up the embryo; the controversies relating to the manner in which the embryo is formed; and the widely different views respecting the origin of both the mesoderm and the entoderm. Kupffer's vesicle still remains a complete mystery; and no one has thus far succeeded in giving a complete and satisfactory account of the origin of the germ-ring ("embryonic rim," Balfour).

No attempt has been made to explain how the alimentary canal is formed; and the precise origin of the chorda and its mode of differ

entiation are questions which have not been exhausted. Some of the general features of the cleavage have been understood since the time of Rusconi; and the researches of Ryder and Hoffmann have demonstrated the existence of polar globules, pronuclei, and karyokinetic figures, in addition to numerous other facts of both special and general importance. But it is manifest that our knowledge in this direction, invaluable as it is, is very far from having reached that degree of completeness with which we are familiar in the case of some other vertebrates, and many invertebrates. The importance of accurate and detailed study of the cleavage phenomena has been illustrated in so many cases in recent years, that it is now fast becoming unnecessary to insist upon it.

Remembering that the histogenetic sundering of the embryonic material actually begins with the cleavage, and that "jeder einzelne Entwicklungsmoment ist die nothwendige Folge des vorausgegangenen und die Bedingung des folgenden," 12 it seems clear what course our investigations should take in order to reach satisfactory conclusions on the origin and relation of the germ-layers. But in all telolecithal vertebrate ova, especially those extreme forms in which the cleavage is restricted to a discoidal mass aggregated at one pole, the difficulties in the way of tracing the precise genealogy of individual cells soon become quite insurmountable. Notwithstanding the exceptional advantages for observation afforded by transparent pelagic fish eggs, no one has hitherto succeeded in tracing the exact genetic relationship of each cell beyond the 16-cell stage.

In passing from the 16-cell to the 32-cell stage, the central portion of the blastodisc becomes two cells deep, and on this account it becomes extremely difficult, beyond the latter stage, to trace the genesis of the individual cells in the living egg. By the aid of mounted preparations we have found it possible to obtain the complete genealogical history of each cell as far as the 64-cell stage. In leaving this stage the blastodisc becomes three cells deep in its central portion, and we have been unable to carry the complete identification of all the cells beyond this point. Fortunately, the more interesting among the concluding phenomena of the cleavage are confined to the marginal cells of the disc; and it is the history of these cells that we have been able to follow with sufficient completeness to decide one of the cardinal questions in the early development of the teleostean fishes,

11 Müller's Arch., 1836, p. 278.

12 Leuckart and Bergmann. Vergl. Anat. u. Phys. d. Thierreiches, p. 19.