MUSEUM OF COMPARATIVE ZOOLOGY. 5 



months later, when I had made microscopic sections of the whole ovarian 

 egg, including membrane and granulosa. 



As the outcome of this journey I secured a number of series of eggs, 

 beginning with the early stages of segmentation, from which I prepared 

 at intervals and by various methods a considerable quantity of material. 

 I was also able to bring to Cambridge about two hundred young fishes 

 just hatched.. Some of these were kept alive until September 26, 1886, 



— nearly three and a half years. 



The fishes brought to Cambridge were put into running hydrant 

 water and fed on mosquito larvae for several weeks, — until about the 

 1st of August, — when they were large enough to swallow small " mud 

 minnows " of nearly their own size. These were gradually substituted 

 for the larvse, and those fishes which were large enough to avail them- 

 selves of this kind of diet grew much more rapidly than their mates. 



I have also received some eggs from Mr. Perry since my last trip to 

 Black Lake, and although his attempts at fertilization did not prove to 

 be more successful than my own, I still hope to secure before long the 

 early stages which are needed to fill the gaps in my material. 



II. Habits of the Young Pishes. 



The habits of young gar-pike have already been quite fully described 

 by A. Agassiz (78 a ) and Wilder ('76, 77), so that I shall not have 

 much to add to what has been previously published. 



When first hatched the fish is so small in comparison with the size 

 of the yolk-sac that it swims only with the greatest difficulty, and its 

 movements are anything but graceful. It is so disinclined to swim, that, 

 were it left alone in water sufficiently pure to meet its requirements, I 

 have no doubt it would not move from the point of its first attachment 

 for many hours, or even days. When hatched in confinement the young 

 fishes always swim nearly up to the surface of the water and attach them- 

 selves to the sides of the dish. When there are a large number of them 

 they may attach themselves to floating objects. Frequently the super- 

 ficial film of the water — aided possibly by secretions from the oral disk 



— serves to support an individual in the middle of the dish. Sometimes 

 half a dozen or more individuals form in a cluster, and appear to han^ 

 suspended simply from the surface of the water. It is evident that they 

 are not merely floating in the vertical position, because in such cases 

 the surface of the water in their vicinity is always more or less depressed, 

 and upon the slightest touch the fishes begin at once to sink slowly ; if 



