Analysis of Heterogeneity in the Population 301 



In the season of 1907 the first generation showed a curious reduction of the 

 line composed of biotypcs 1, 2, 3, and 4; 1 and 2 were entirely absent in the 



12-|^ 4 29 31 W-r'iJl^A 9 21 44 



42-:*iK% I j ' 46 -V^|$, , 1 i' I 



17-,-liir'- i i / .W-7 81 



46 aft ^' ' ^V"--42 42 m '■■'-'^' ^^''--^ 



«, ^ «Wf/ 1:8;^— -63 21 -^ <«Jt;^^||i^_..77 



9 ^ <^«te^^.— .74 ' ^ ; ^«»^.__..8i 



45 jm^ ^, jj^ 94 -^< ^..____..88 



47 ^ ^ - __i4g 109 ■-'^^■mt^-r-^..~-m 



51 )eySi 4»> 219 108 '-4ISfl»' iS^-- 196 



iW* i^ 314 , ^ 431 



^/iB^ -«^- 198 '^.'^, ^--^-----219 



/ \ 'V? '- -- 85 



/ V-W- 9 \ 



14 19 " 41 



46 



1955 Males. Total 4221. 22G6 Females. 



Fig. 105. — Census of first annual generation at Puebla In 1006, showing same 

 array of pattern conditions present in figure 104, but somewhat decreased in 

 amount of fluctuating variability. 



1787 Males. Total 3695. 1908 Females. 



Fig. 106. — Census of second annual generation at Puebla in 1906, showing a 

 restriction in the array of pattern conditions, the appearance of isolated groups 

 and increased fluctuation in the line composed of biotypes 1, 2, 3, and 4. 



males, while 1 was present and 2 and 3 absent in the females. The same condi- 

 tion prevailed in the second generation of this year, where biotype 1 was present 

 in both sexes, but 2, 3, and most of 4 were absent, as were also 10, 11, and 12 in 



