MOETALITY IIS^ PIKE-PEKCH EGGS IX HATCHERIES. 3 



therefore, comes on that part of the protoplasmic investment of tlie yoke which 

 covers the oil globule and here it bursts. In almost every case the white spot 

 which indicates the rupture of the yoke investment makes its appearance at 

 the oil globule, usually at the equator. 



Almy's as well as our own observations show that the death rate 

 increases rapidly and steadily to the fourth day and then advances 

 more slowly. To begin with, it must be noticed that a smalT per- 

 centage of dead eggs is found practically as soon as the fishes are 



Age oj- egos itl rtours. 



Fig. 1. — Graph .showing tlie variation in number.s of abnormal and unscgmentpd pike- 

 perch eggs during development. 



stripped. It is surmised that these may haye been injured in the 

 process of stripping, or that they may haye died through some de- 

 velopmental irregularity while still in the fish. 



Coming now to the hypothesis that failure of fertilization is re- 

 sponsible for a greater part of the mortality of hatching eggs, it is 

 generally assumed that lack of impregnation and failure to segment 

 are closely correlated. A detailed examination of the material does 

 not bear this out. We found that in eggs 4 hours 30 minutes old there 

 was a considerable percentage which showed no trace of cleavage. 

 This was true also at 5 hours 30 minutes, 6 hours 45 minutes, and 



