90 CORRELATED VARIATIONS. 



No. of Segments, 234 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 

 Frequency, 10 7 20 303 412 534 1552 223 59 35 43 6 6 2 



In this plant Pearson found the 8 segment capsules 

 to be highly fertile, whilst in the 10, 11, and 12 segment 

 capsules he could find hardly any seed at all. Six and 

 7 segment capsules were only moderately fertile, and 

 those with 5 segments or less were practically sterile. 

 These experiments Professor Pearson holds to illustrate 

 a very important law, namely, " Fertility is not uni- 

 formly distributed among all individuals, but for stable 

 races there is a strong tendency for the character of 

 maximum fertility to become one with the character 

 which is the type." It follows, therefore, if this prin- 

 ciple is generally true, that stable races are very largely 

 the product of the typical or most frequently occurring 

 members, and not of all the individual members in pro- 

 portion to their numbers. There would seem to be a 

 constant tendency to keep the type uniform and limit 

 the variability in either direction as much as possible. 

 Doubtless under changed conditions of environment, 

 the relative fertility might also become changed, and in 

 consequence a gradual evolution result. 



A similar relationship between fertility and type 

 form has been noticed by Davenport * in one of the 

 Hydromedusse, Pseudodytia pentata. This organism 

 differs from all other Hydromedusse in that it normally 

 has five radial canals, instead of four. A. G. Mayer f 

 has examined the variation of the species, and obtained 

 the frequencies given in the subjoined table. From 

 these values we gather that the four and six canal forms 



*Biometrika, i. p. 255, 1902. 



\ Science Bulletin of the Brooklyn Museum, vol. i. 



