July 22, 1910] 



SCIENCE 



121 



WX — W (male) 

 RX— RX (female) 



RWXX (50^) — RWX (50^) 

 Red female Red male 



Wien these F^ individuals are mated, the 

 following table shows the expected combina- 

 tions that resiilt : 



RX — WX (Fi female) 

 RX— W (Fi male) 



RRXX— RWXX — RWX— W WX 



(25^) (25^) (25^) (25^) 



Bed Red Red White 



female female male male 



It will be seen from tjie last formulas that 

 the outcome is Mendelian in the sense that 

 there are three reds to one white. But it is 

 also apparent that all of the whites are con- 

 fined to the male sex. 



It will also be noted that there are two 

 classes of red females — one pure REXX and 

 one hybrid RWXX — but only one class of red 

 males (RWX). This point will be taken up 

 later. In order to obtain these results it is 

 necessary to assume, as in the last ■ scheme, 

 that, when the two classes of the spermatozoa 

 are formed in the F^ red male (RWX), E and 

 X go together — otherwise the results will not 

 foUow (with the symbolism here used). This 

 all-important point can not be fully discussed 

 in this communication. 



The hypothesis just utilized to explain 

 these results first obtained can be tested in 

 several ways. 



Verification of Hypothesis 

 First Verification. — If the symbol for the 

 white male is VV WX, and for the white female 

 WWXX, the germ cells will be WX— W 

 (male) and WX — WX (female), respectively. 

 Mated, these individuals should give 



WX— W (male) 

 WX— WX (female) 



WWXX (50^) — WWX (bOi) 

 White female White male 



All of the ofl'spring should be white, and 

 male and female in equal numbers; this in 

 fact is the case. 



Second Verification. — As stated, there 

 should be two classes of females in the F, 



generation, namely, RRXX and RWXX. 

 This can be tested by pairing individual fe- 

 males with white males. In the one instance 

 (RRXX) all the offspring should be red— 



RX— RX (female) 



WX — W (male) 



RWXX — RWX 



and in the other instance (RWXX) there 

 should be four classes of individuals in equal 

 numbers, thus: 



RX — WX (female) 



WX— W (male) 



RWXX— WWXX — RWX— W WX 



Tests of the F^, red females show in fact 

 that these two classes exist. 



Third Verification. — The red F^ females 

 should all be RWXX, and should give with 

 any white male the four combinations last 

 described. Such in fact is found to be the 

 case. 



Fourth Verification. — The red F^ males 

 (RWX) should also be heterozygous. Crossed 

 with white females (WWXX) all the female 

 offspring should be red-eyed, and all the male 

 offspring white-eyed, thus: 



RX — W (red male) 

 WX — WX (white female) 



RWXX— WWX 



Here again the anticipation was verified, 

 for all of the females were red-eyed and all of 

 the males were white-eyed. 



Crossing the New Type with Wild Males and 

 Females 

 A most surprising fact appeared when a 

 white-eyed female was paired to a wild, red- 

 eyed male, i. e., to an individual of an unre- 

 lated stock. The anticipation was that wild 

 males and females alike carry the factor for 

 red eyes, but the experiments showed that all 

 wild males are heterozygous for red eyes, and 

 that all the wild females are homozygous. 

 Thus when the white-eyed female is crossed 

 with a wild red-eyed male, aU of the female 

 offspring are red-eyed, and all of the male off- 

 spring white-eyed. The results can be ac- 

 counted for on the assumption that the wild 

 male is RWX. Thus: 



