[ No. 64] DIPTERA OF CONNECTICUT: MORPHOLOGY 27 



rated, as the eye-stalks on which they are borne become more and more 

 lengthened. In the male of Richardia telescopica^ however, the an- 

 tennae remain close together, while the eye-stalks become greatly pro- 

 longed (as a sexual character exhibited bj^ the male alone). 



The antennae may exhibit some very striking sexual differences 

 in the Diptera. Thus the segments of the antennae of the male of the 

 tipulid Gerozodia plumosa bear long processes which give a feather}' 

 appearance to the antennae (compare Fig. 5, U) in this sex, although 

 the antennae of the sub-apterous female of this species are of the 

 normal type. A more familiar type of sexual dimorphism in the an- 

 tennae is illustrated by the densely verticillate '"feathery" antenna of 

 a male mosquito, such as the one shown in Fig. 5, V, which presents a 

 marked contrast to the sparcely verticillate type of antenna occurring 

 in the female (Fig. 3, E). Among the higher Diptera, the tachinid 

 fly Talarocera nigt'ipenyiis, figured by Williston (1908, p. 31), has the 

 third segment of the antenna of the male modified in the peculiar 

 fashion shown in Fig. 5, R, while the third segment of the antenna of 

 the female is rather deeply cleft, giving it a somewhat furcate ap- 

 pearance. 



The antennae of the males of such tipulids as Megistocera fZipes, 

 Macroiiuistix costalis^ etc., are extremely elongated, although the num- 

 ber of antennal segments is reduced to about eight, and the length- 

 ening of the antenna (which may be one or more times the length of 

 the body) takes place through the lengthening of the individual seg- 

 ments, rather than by an increase in the number of the segments. In 

 I the mycetophilid genus Macrocera^ the antennae are extremely elon- 

 gated in both sexes. 



The antennae are extremely reduced in the Hippoboscidae ; and 

 the segments are so greatly modified that Dufour (1845) thought they 

 consist of but a single immobile segment in these flies. Traces of 

 three segments may be found in the antennae of the hippoboscids. 

 however, although the first segment is so reduced and so highly modi- 

 fied that it is difficult to recognize it. The second segment, labelled p 

 in Fig. 5, 0, is the largest, and conceals the third segment, which is 

 withdrawn into the sacciform second segment (from which the arista, 

 «r, borne by the third segment, may be protruded). 



According to Felt (1921), "The normal number of antennal seg- 

 ments among generalized Nematocera is probably 16 — ^that is, a 

 greater or smaller nmnber means specialization by addition or reduc- 

 tion." Felt (1925) likewise records the remarkably great number of 

 41 segments* for the antenna of the cecidomyid Lasioptera perarticu- 

 lafa, while the Mexican cecidomyid Ceratomijia, described by him, 

 has the antennae reduced to "comparatively insignificant and presum- 

 ably functionless organs with but six segments" (which is the mini- 

 mum for the Nematocera in general), so that the cecidomyids present 

 as great a variation in the number of antennal siegpients as is to be 



♦Alexander (1936) has described a cecidomyid from Panama, Feltomyia polymera 

 (chang-ed to Feltomyina polymera hv Alexander, 1937), having 65 antennal segments; 

 and Tonnoir (1939) has described an African bruchomyine psychodid, Bruchomyia 

 edwardsi (changed to Eutonnoiria edwardsi by Alexander, 1940), having the remark- 

 ably large number of 113 antennal segments— the greatest number of antennal seg- 

 ments thus far recorded for the Diptera. 



