32 



As it seemed likely that for some reason a large per cent. 

 of the eggs failed to be impregnated, my attention was 

 first directed to determining the iirst differences between 

 impregnated and unimpregnated eggs. 



In order to determine the qnestion witli certainty for 



this particnlar animal, the following experinjent was 



tried (quoted from note-book) : 



April 1G, 8:45 a.m. 



After washing the surface of the body of a female hsh 

 in the region of the external opening with weak acetic 

 acid, in order to destroy any spermatozoa, the eggs were 

 stripped into dishes containing water. Into one dish milt 

 was immediately stripped ; the other was left without milt. 

 These were marked lot 1 and lot 2 resi3ectively. 



Lot 1. Examined at 1:45 p.m. (the eggs having been kej^t 

 in a cold room), and found segmentation going on. The 

 germinal disk is divided into either two or four cells. 



Lot 2. Examined 8 a.m., April 17th, twenty-four hours 

 after impregnating lot 1. The eggs were lirmly set in a 

 mass on the bottom of the dish. One hundred taken at 

 random w^ere examined witli following results : 



Showing normal germinal disk without trace 



of segmentation 82= 82;^ 



Showing abnormal germinal disk with possible 



traces of first or second segmentation 4= 4^ 



Injured by rupture of protoplasmic invest- 

 ment of yolk, so that yolk had escaped 

 and egg had turned white 14= 14^ 



Total 100=100^ 



This experiment was afterwards repeated without, how- 

 ever, counting the eggs, and always with the same result. 

 Segmentation of the germinal disk is, therefore, the first 

 easily recognized sign of impregnation. 



In order to determine the percentage of unimin-egnated 

 eggs among those taken by the men and ready for ship- 



