324 Genetics of a Daphnia Hybrid during Parthenogenesis 



among the females of the hybrid clone, the original hybrid female was 

 omitted, and the correlation between mother and daughter for the 



-2 ratio was calculated for the population formed by the 16 of the 17 lines 



which left female offspring. All generations are combined in the table, 

 and therefore many specimens are counted twice, once as offspring of the 

 wth generation, and again as parents of the (n + l)th generation. When- 

 ever a parent had more than one offspring, it is counted over again with 

 each one, in the usual way. Thus the number of offspring available for 

 the correlation was 276^ but the actual number of parents is only 77. 



The crude correlation table yields a positive correlation coefficient 

 of '132 + '039. This, being more than three times its P.E. must be con- 

 sidered as probably significant. This is, of course, far from equivalent to 

 saying that it is due to genetic resemblance between mother and 

 daughter. Indeed, detailed examination shows that it is almost certainly 

 to be ascribed to environmental causes. 



The readiness with which somatic modifications are induced in Clado- 

 cera by environmental changes is well known, having been studied by 

 many writers, including the author. Especially significant for our present 

 purposes are the large parental correlations, both positive and negative, 

 which may be produced within a clone by changing environment (1914). 

 In order to discover if there were &ny prima facie justification for sus- 

 pecting the operation of changing environment (food, temperature, etc.) 

 in producing the parental correlation just described, the six months 

 during which the hybrid clone was under observation were divided into 

 nine periods of three weeks each (roughly the period covered by one 

 parthenogenetic cycle), and the mean ratios for all females born in each 

 period was determined, with the following result : 



Period 



1. April 19— May 9 



2. May 10— May 30 



3. May 31— Jane 20 



4. June 21— July 11 



5. July 12 — August 1 



6. August 2 — August 22 



7. August 23— September 12 



8. September 13 — October 3 



9. October 3— October 24 



The fact that the means for the last four 3-week periods are all 

 lower than those of any of the first five, obviously tends to produce a 



1 Two females of the clone were omitted from this, and all other calculations, owing to- 

 abdominal deformity. 



