1098 BULLETIN OK THE BUREAU OF FISHERIES. 



Some older records for the Woods Hole region are as follows: Storer(i855) 



found eggs or young in June, July, and August; Goode (1884), at Noank, Conn., in 



1874, founds eggs on July 14, and young one-half inch long on July 21, which 



had, on September i, reached a length of i inch. Ryder (1886) says: 



Oviposition occurs about the middle of July in the latitude of Woods Hole. How 

 long it lasts has not been determined, but judging from the condition of the roes and 

 milt of the adults at that time, it seems probably that they do not spawn later. 



According to Bumpus (1898) : 



Oviposition occurs as early as June 3, and it may occur at any subsequent time 

 throughout the month. 



The higher temperature at Beaufort, however, allows the breeding season 

 to be long drawn out. I have found advanced embryos the first week in June and 

 eggs in segmentation late in August. Thus the spreading out of the time of 

 oviposition makes it easy to find spent females and correspondingly hard to 

 find ripe ones. I may add that no other ripe females have been found this 

 season (1908). 



Formation of the germ disk. — The germ disk was quite definitely formed 

 as a fairly distinct white patch about three hours after impregnation. The 

 streaming of the protoplasm to form this was in part preceded and in part 

 accompanied by the collecting of yellow oil globules at the micropylar region. 

 These oil drops, some large and some small, form in a layer under and visible 

 through the germ disk in just the same fashion as the writer found in the egg 

 of the pipefish, and as has been reported for a large number of teleosts. For a 

 discussion of the matter see Gudger (1905). Miss Wallace (1898) also reports 

 oil globules in the egg of this fish. The streaming of the protoplasm continues 

 after the segmentation has begun, thus the germinal disk grows constantly larger 

 for awhile by these additions of protoplasm. Ryder's (1886 and 1887) error in 

 saying that there are no oil drops in the eggs of the toadfish is hard to explain 

 save on the supposition that they disappear by the time those stages on which 

 he worked are reached. 



Segmentation. — As already stated, segmentation begins seven hours after 

 fertilization at Woods Hole, five hours after that act at Beaufort. The ger- 

 minal disk divides in a vertical plane into two cells, then along another plane, 

 at right angles to the second, into eight cells. The segmentation into two and 

 four cells results in blastomeres about equal in size and symmetrical in place. 

 But beginning with eight cells and continuing on until division is ended, the 

 segmentation is very irregular, growing more so with each division. 



Notwithstanding the large size of the yolk, the blastoderm never piles up 

 high with a clearly defined circumference as is found in the Salmonidse, whose 

 eggs are about the same size, and as the writer found to be the case in the pipe- 

 fish (Gudger, 1905), whose eggs are only one-fifth (i mm.) as large. On the 



