The Process of Heredity 513 



learn to see the outlines of fixity and regularity through the haze of 

 a confusing diversity of conditions. 



No two strains develop at the same rate, but are retarded or 

 accelerated by various conditions of temperature, oxygen content 

 of the water, etc.; only when a pure and a hybrid strain are started 

 simultaneously and are treated alike, can any basis of comparison 

 be attained. 



Even in pure strains, moreover, there is a considerable amount 

 of variability; imperfect eggs are always present and give imper- 

 fect and unhealthy embryos. The great majority of eggs in pure 

 strains, however, especially in the case of F. heteroclitus, develop 

 at a practically uniform rate for the first week. There is always 

 a considerable degree of variability in respect to the hatching 

 period, some individuals hatching as much as four days later than 

 others. 



In hybrids of both kinds the factor of variability is a much more 

 important one than in the case of the pure strains. Here the 

 range of variability is enormously increased and it is a matter of 

 considerable difficulty, especially in advanced stages, to decide on 

 the average or representative condition when a hundred or more 

 embryos are examined. Considerable practice, however, has 

 made it possible, with some degree of personal satisfaction, to 

 select a specimen that represents either the most prevalent condition 

 or a judicious mean between extreme variants. In some cases, 

 especially during the first four days of development, the selection of 

 a representative condition offers no serious difficulties. It must be 

 understood, then, that, in the description and figuring of such 

 representative individuals in the succeeding pages, these have been 

 arbitrarily selected by the writer and that his best judgment has 

 been used in the selection. Where actual measurements or numer- 

 ical determinations of structures or functional activities could be 

 employed there was no difficulty in selecting the most prevalent or 

 the average condition. 



A fuller discussion of this factor of variability, especially as it 

 applies to hybrids, follows the presentation of data. 



