GENE MUTATIONS 



but rather weak. Antennal segments are length- 

 ened but are not as attenuated as are those of 

 long (Fig. 8). Leg segments are also somewhat 

 lengthened but heavier than corresponding seg- 

 ments of long. Distinction of semilong as of 

 long from wild-type is clear in antennae and to 

 a great extent in legs, but overlapping between 

 long and semilong apparently takes place espe- 

 cially in legs. The wing of semilong is quite 

 flat. Distal shortening, however, takes place 

 as is evidenced by the form of the outer margin 

 and especially by the abbreviation of the first 

 radial cell. A slight indentation is present 

 at the apex of the wing, not however, as pro- 

 nounced as in fused. It was first thought that, 

 semilong might be allelic to long, but this) 

 proves not to be the case. 



TAPERING, ta (antennae). In May, 1930, nine 

 males with tapering antennae v/ere found in a 

 wild-type culture. The antennae are very defi- 

 cient with much fusion and irregularity of seg- 

 ments distally (Fig. 8). Tapering does not 

 overlap with type, but reverse mutations have 

 occurred (P. W. Whiting, 1932a). Both sexes^ 

 are fully fertile and of normal viability. 



ANTENNAPEDIA, ap (antennae). In the course of 

 genetic work involving tapering, v/asps were ob- 

 served with tarsal-like claws on the terminal 

 segments of the antennae. A more or less well- 

 formed foot may be present (P. W. Whiting, 

 1934b) . The distal segments may be modified to 

 resemble a tarsus v/ith the terminal segment 

 provided with claws, arolium, and calcanea as 

 in the normal case. The foot may be present on 



57 



