210 ROSCOE R. HYDE 



females; a ratio of 1 crescent to 40.3 longs. I shall neglect this 

 ratio in further discussion as it appears to be aberrant. 



In the cross as given in table 6 a, there were 34 crescent males 

 and 19 crescent females, a ratio of 53 crescent to 1127 long or 1 to 

 21.3. In the reciprocal cross, table 6b, there were 30 cres- 

 cent males and 49 crescent females in a total of 1714. This 

 gives a ratio of 79 crescent to 1635 long or 1 to 20.7. 



The different classes appeared in Fo generation of the crosses as 

 follows. From a total of 16098 as given in the cross in table 9 

 there are 516 truncate males, 778 truncate females, 431 crescent 

 males, 633 crescent females, 6881 long males and 6857 long females. 

 In the reciprocal cross there are 273 truncate males, 332 truncate 

 females, 315 crescent males and 436 crescent females, 4986 long 

 males and 5458 long females. Out of a total of 27896 there are 

 1903 truncates and 1881 crescents. This gives a ratio of 1 trun- 

 cate to 13.2 longs and 1 crescent to 13.4 longs. The ratio is 

 practically the same in both the cross and the reciprocal. To 

 summarize in a general way, we may say that during the period 

 of investigation the truncates threw one long in 7 =t truncates. 

 Their long-winged brothers and sisters threw 1 truncate in 7± 

 longs. The new type of wing, the crescent, appeared in the crosses 

 between the truncates and the wild stocks and in the ratio of 1 

 crescent to 21 longs in both the cross and its reciprocal. In the Fo 

 generation crescents and truncates appear in equal numbers from 

 both the cross and the reciprocal and in the ratio of 1 to 14. 



These ratios apparently do not fall under any Mendelian 

 explanation and yet I believe that any explanation of these ratios 

 must take into consideration the great viability of the truncate 

 stock and my opinion is that when more facts are available this 

 case may be found to fall under a Mendelian formula because 

 there is evidence of segregation and because definite ratios appear 

 according to the combinations that are made. 



