INITIATION OF DEVELOPMENT IN CH^TOPTERUS. 27 



through cheese-cloth to free them from body fragments and from 

 some of the mucus, which is very thick. 1 



The eggs were left in sea-water, after removing from the 

 ovaries, from twenty to forty minutes, in order that the first 

 maturation spindle might form before further treatment. Ex- 

 periments showed that eggs so treated formed a larger percentage 

 of second polar bodies. Such eggs formed the polar bodies more 

 quickly, both in normal fertilization and in experiments. Un- 

 fertilized controls were always run, and fertilized controls also 

 if the experiment included the effect of the combination of 

 fertilization and artificial parthenogenesis. Care was taken, of 

 instruments and dishes, to keep all unfertilized eggs free from 

 contamination from sperm. 



A considerable amount of material was fixed for cytolog\cal 

 study, largely covering the ground of the experiments. The 

 study of the fixed material was used to amplify and check that of 

 the living eggs. The eggs used for this work were killed either 

 in Boveri's picro-acetic; in Flemming's (weak) solution; or in 

 Meves's fluid. They were stained in Heidenhain's iron-haema- 

 toxylin or in thionin, and counter-stained in Orange G. The 

 sections were cut four micra thick. 



II. THE FIRST CHANGES IN DEVELOPMENT. 



So long as the eggs remain in the ovary the germinal vesicle is 

 intact, but almost immediately upon entrance into sea- water 

 the vesicle breaks down, its contents migrate to the animal pole, 

 and the metaphase of the first maturation spindle forms. At the 

 same time the membrane, which could be seen from the beginning, 

 stands off from the egg at a greater distance, forming a distinct 

 space between the membrane and the egg surface. Here de- 

 velopment pauses, and unless further stimulus is given the eggs 

 remain in this condition, and after a number of hours go to 

 pieces. 



The question of determining the factors which bring about 

 these first developmental changes is a problem of which I have had 



1 It is probable that one reason so many of the controls died before the close 

 of the experiment was that more mucus was left in the control dishes than in the 

 ones in which the solutions were changed a greater number of times. 



