THE ORIGIN OF GYNANDROMORPHS. 77 



as striking as these are found in other parts of the body and come out 

 equally well in the gynandromorph. Miss Mehling shows that the 

 male and female parts may sometimes, however, be so intimately 

 combined that a particular organ, such as a leg, may seem, on super- 

 ficial examination, to be a blend of the two. A minute examination 

 shows, however, as a rule, that such an organ is a piecework or mosaic 

 of male and female characters. 



It will be recalled that Boveri's hypothesis appealed to the phenom- 

 enon of partial fertilization. A belated sperm, sometimes failing to 

 fuse with the egg nucleus before the latter divides, comes to combine 

 with only one half of the latter. As a result, one of the first two 

 segmentation nuclei contains only the maternal daughter nucleus, 

 the other the combined maternal daughter nucleus and the entire 

 sperm nucleus. The application of the results to the gynandromorph 

 bees is obvious. If these are due to partial fertilization, then we 

 should expect the male side to be like that of the mother's race the 

 Italian bee because its haploid chromosome group came directly 

 from the Italian mother's egg. Vice versa, the female side should 

 show hybrid characters, or the Italian character if the Italian race 

 dominates completely the German race. If the latter, both sides 

 would then be alike and racially indistinguishable. Morgan's sug- 

 gestion of polyspermy leads to the following explanation: If under 

 unusual circumstances one (or more) of the spermatozoa should 

 develop, the parts supplied by its nuclei would be haploid, hence male, 

 while the other parts resulting from the combined nuclei would be 

 female. The expected characters of the two parts of the gynandro- 

 morph would be the reverse of those called for by Boveri's hypothesis, 

 for the male parts should be paternal on Morgan's view, maternal 

 on Boveri's. The decision lies, therefore, in the character of the male 

 parts of the gynandromorph. Boveri examined von Siebold's bees, 

 some of which had been preserved in Munich, in order to get an 

 answer to this problem, and reached the conclusion that the male 

 parts are maternal. Hence the answer was in favor of his own hypo- 

 thesis. We may now proceed to examine this evidence in detail and 

 then see whether the hypothesis of chromosome elimination may not 

 fit the facts as well as either of the alternative views. 



After nearly 50 years in alcohol the Eugster gynandromorphs had 

 lost so much of their color that a comparison with the racial pigmenta- 

 tion as seen in the living bees was impossible. Only after extracting 

 the superficial pigment and dissolving away the external parts of fresh 

 individuals of the two parental races was it possible for Boveri to 

 make any reliable comparisons. Even then only a few individuals 

 were available, because "an vielen Exemplaren das Abdomen nahezu 

 farblos ist." Boveri confines his account to four specimens and in 

 these takes only the head and abdomen into account. 



