( 507 ) 



membrane. This tunnel is large in snmc genera (^Anii/role/iidid, for instance), 

 and (Mmparatively small in others. It is rntlimentary among Noctiiidai'. 



(3) The ttimilies in which the cavity lies nnderneath the pleura of the second 

 abdominal segment ; the pleura of the first segment is small and longitudinal, being 

 placed above the cavity. This is the case, for instance, in Geometridae, Uraniidae, 

 and Pi/ralidac. 



I shall have to refer in another place to the taxonomic value of the organ 

 under discussion. However, one set of characters I think deserves special mention. 

 The organ itself is not glandular, but has in several families been taken into the 

 service of glands restricted to the male se.x. Such a secondary sexual character 

 occurs widely among Geometridac (to which family I confine my remarks) in 

 connection with the development of a scent-organ situated in the hindtibia. This 

 tibia is swollen in the males of many species, bearing on the inuerside a deep 

 slit, from which projects a brush of long stiff hairs wlien the slit is open. The 

 brush is the distributor of the scent produced by special glands in the tibia, 

 such a brush of hairs being commonly found in Lepidoptera in connection with 

 scent-producing glands. Now, in these males the npper proximal angle of the 



Explanation of Figuke. 

 MR = me.<onotum; MT = metanotiun ; C = hiadco.x.a ; P'= pleura of firist. abdominal se^^niont. bearing 



the first .stigma ; F- 

 sesment. 



r pleura of seeond segment, bearing the scconcl stigma ; S^ = sternitc of second 



sternite of the second abdominal segment (Fig., R-) is produced into a sjiino-like 

 process which projects free over the cavity of the sensory organ. Tlie hindtibia of 

 the insect lies against this process, and from the position of the scent-brush there 

 can hardly be any donbt that the process is employed to spread the brush out 

 l)y the tibia being rubbed against it. In fact, when holding a live Bo'irmia by 

 tiie wings, the working of the hindtibia against the spine and the spreading of 

 the brush can be observed. The process occurs apparently only in Geometridae. 

 It is sometimes long, sometimes short, and may be strongly chitinised or may be 

 very weak. One finds such different ]ihyletic stages of develojiment in closely 

 allied species, and it appears to me probable that very often in near allies the 

 process has been lost with the reduction or loss of the tibial scent-organ in some 

 species, and preserved together with the scent-organ in other species. Another 

 i|uestion naturally presents itself: docs the process occur in all the species of 

 Geomrtridae of which tlie hindtibia beacs a scent-organ in the male 'f So far, \ 



