EXPERIMENTAL EVOLUTION. 



127 



Table 8. — Reactions of mixtures of motise and rat s-perm-cells to acid solutions. 



From the above data it may be concluded that there is differential 

 mortalit}^ in the sperm-cells of different species when subjected to weak 

 solutions of acetic acid. 



SELECTION OF SEX INTERGRADES IN DAPHNIA. 



As noted in last year's report, Dr. Banta has undertaken to increase 

 or decrease the amount of a sex-intergrade condition that he dis- 

 covered in the water-flea, Daphnia, by means of selection based on 

 somatic differences. At that time it was thought that it could be con- 

 cluded that strains like No. I, selected for increased intergradeness, 

 would probably not become more intergrade; but strains Hke Nos. Ill 

 and V, selected to produce normal females, could tend to become so. 

 All mothers received identical treatment so far as practicable — culture 

 water was taken from the same jar and the temperature was maintained 

 in each at the same level. By means of a somewhat arbitrary scale, 

 running from to 80, the degree of intergradeness may be expressed 

 quantitatively. It is to be recalled that all strains descended partheno- 

 genetically from the same progenitor. 



The procedure and the results with intergrade strains I and III were 

 as follows : Strain I was maintained as a "high " intergrade strain. Its 

 level fluctuated from 10 to 33, averaging about 22 on the scale of inter- 

 gradeness. Strain III was selected as a ''low" strain. Selection 

 was effective in the seventh generation, the level falling to about 3. 

 After five further generations of selection the level fell to about 1 . In 

 the nineteenth generation there was an abrupt rise in this low strain to 

 20, after which, during four generations of selection, it fell to 0, produc- 

 ing none or very few slight intergrades for the next five generations. 

 Later, it again went to somewhat higher levels. However, in spite 

 of all the fluctuations in both the high (I) and the low (III) strains 

 after the sixth generation a pronounced divergence between the two 

 strains was maintained (fig. 2). 



In the nineteenth generation a return selection was begun in the low 

 strain (III) to produce a high strain. Unfortunately this was begun 

 when Strain III was at a rather high level, but the result was the same 

 in Strain V, another low strain, which was at a low level when return 



