Tlie Process of Heredity 511 



Not only nuist one be careful as to the time for starting experi- 

 ments, but individuals must be carefully chosen. Males and 

 females, full grown and sexually mature, as indicated by the sure 

 signs of the sexual climax mentioned above, should always be used, 

 for if eggs are stripped from females that are under-ripe only a 

 small per cent of the eggs are capable of fertilization and the devel- 

 opment of those is apt to stop short of completion. If males 

 that have either failed to reach or have passed the sexual climax 

 are used, their milt may be either entirely or largely ineffective 

 in initiating normal development. 



The eggs and milt of fish kept longer than three or four days in 

 aquaria very frequently become stale, and very unsatisfactory 

 results have been obtained from their use. These stale eggs are 

 frequently capable of fertilization and of partial development, but 

 the embryos usually die before hatching. This precaution applies 

 especially to F. majalis, which very seldom spawns in captivity 

 and hence females may carry eggs in the oviducts for weeks after 

 they would normally be extruded. F. heteroclitus females, on 

 the other hand, have a habit of ridding themselves of over-ripe 

 eggs without the assistance of the males. Consequently stale 

 eggs are seldom found in that species. Stale milt may be avoided 

 in both species by selecting only the most highly colored males, 

 with well developed contact organs. F. majalis males will not 

 retain their dusky spawning coloration for more than a few days 

 in captivity, so there is little danger of obtaining stale milt from 

 that source if reasonable care is taken. 



A multiplicity of parentage in a single batch of eggs must be 

 avoided, as there is considerable variability in the eggs taken from 

 different females and in the milt of different males according to 

 their size, age, and degree of sexual maturity. Unless the eggs 

 of one female are used for both pure and hybrid strains, and the 

 milt of one male for both species of eggs in one series of experi- 

 ments, an undue amount of variability ensues as the result of 

 extraneous factors that needlessly complicate the issue and frus- 

 trate all attempts to study pure heredity. 



