146 H. H. Nezuman 



to be fertilized or are incapable of development. A few might 

 develop if allowed more time. No attempt, however, has been 

 made to distinguish between unfertilized and retarded eggs, as 

 cytological methods would be required. 



In all experiments every precaution was taken to treat the lots 

 of eggs used for pure and hybrid strams exactly alike. In some 

 of the experiments a mixed lot of eggs from several females and the 

 mixed milt of several males of each species yielded good results, 

 while in other cases the eggs of one selected female were divided 

 and fertilized by the milt of two or more selected males of each 

 species. In every instance the eggs after stripping were thoroughly 

 mixed by stirring and shaking and then divided into two approxi- 

 mately equal lots, which were fertilized at the same instant by 

 abundant milt obtained by macerating the ripe testes of selected 

 males. When development had proceeded to the desired point 

 the two lots, pure and hybrid, were killed at the same instant in 

 equal amounts of picro-sulphuric acid. It was found advanta- 

 geous to examine and count the eggs while in this killing solution 

 because the clear definition of the blastomeres is lost if the eggs 

 are transferred to alcohol. To obtain the best results the examina- 

 tion and enumeration of eggs should be made within a few days 

 after killing. 



After the counts have been made there are obviously several 

 methods of dealing with the data thus obtained. The best method 

 involves a more or less arbitrary valuation of the various stages 

 in terms of blastomeres, the uncleaved and unripe eggs being 

 omitted from consideration. The total number of blastomeres 

 divided by the number of eggs will give the average condition of 

 the lot and will furnish a numerical comparison between the 

 stages of advancement of the pure and hybrid strains. A simpler 

 method consists of a mere comparison of the relative percentage, 

 in the two lots, of more and less advanced conditions, the most 

 significant comparison being that between all 4-cell stages (includ- 

 ing 4-minus) and all 2-cell stages (including 2-plus and 2-minus), 

 3-cell condition being ignored because not strictly assignable to 

 either class. Another simple method involves a comparison of 

 4-cell, 2-cell and intermediate stages in the two strains. A full 



