MUSEUM OF COMPARATIVE ZOOLOGY. 241 



Four hours after an Agalma was placed in the aquarium, eggs in the 

 4-cell stage were picked out of the water in which it was confined. I 

 have traced one and the same egg from the 2-cell to the 4-cell stage, and 

 find that it takes 2 h. 10 m. for the necessary changes to be perfected in 

 this growth. On another egg it was determined that it takes 45 m. to 

 develop an egg in the 2-cell stage from an egg in which the germinative 

 vesicle, or "nucleus," had disappeared. By this observation it will be 

 seen that it requires a little over an hour to pass from the egg just fertil- 

 ized into the stage which exhibits the firt^it sign of a primitive cleavage, 

 plus the interval of time which elapses after the 2-cell stage is formed 

 and before it begins to form the secondary furrow, or origin of the second 

 cleavage-plane. This last interval is probably not more than 30 m. ; 

 consequently the interval which elapses after fertilization before the 

 formation of the primary furrow is about half an hour. 



Impregnation probably takes place in the gonophore. I have not 

 been able to fecundate the Agalma egg artificially, nor was it seen to 

 take place naturally. I have repeatedly tried to fertilize ova with sperm 

 from the same colony, but have always failed. This fact led me, in 1880, 

 to state that the animal cannot be impregnated by spermatozoa from its 

 own male bells. Last summer (1884), however, to obtain some informa- 

 tion on this point, an isolated Agalma was kept in a glass jar, and it 

 dropped eggs which became segmented and later developed into primitive 

 larvae. The water in which it was confined was not changed meanwhile, 

 nor new liquid added. Of course this experiment does not absolutely 

 demonstrate that the spermatozoa from the same colony can or cannot 

 unite with an unfecundated ovum of the same, for sperm may have been 

 in the water before the animal was placed there. Experimentation on 

 the subject has many difficulties ; but it must be confessed, that, as 

 far as I have thus far gone in my studies, it looks as if the male bells 

 of an Agalma may sometimes fertilize ova from the same axis. The 

 great difficulty in the artificial fecundation of the Agalma egg was 

 pointed out by Metschuikoff".* The ovum in the gonophore is enclosed 

 in what he calls an " Umhiillung," from the walls of which the tender 

 egg cannot be extracted without harm to its contents. 



The first naturalist to fertilize artificially the Siphonophore egg was 

 Gegenbaur.f Metschnikoff't was equally unsuccessful with myself with 



* Studien liber die Entwickeiuiig aer Medusen und Siphonophoren. Zcit. f. 

 Wiss. ZooL, XXIV. p. 49. 



t Beitrage zur naheren Keutniss der Schwimpolypen (Siphonophoren), p. 49. 

 t Op. cit., p. 49. 



VOL. XI. — NO. 11. 16 



