Séance du 7 Décembre 1921 123 



length of the antenna should be measured the da> 

 after the moult and again a few months later on the 

 same specimen kept alive for the purpose. 



The study of figs. A and T) has given us some 

 information as to how the aniennal teeth arise, bui 

 has taught us nothing concerning the essential point: 

 the increase in the number of the joints. The best 

 way of filling the gap is to examine preparations of 

 various stages. Let us first try to get a few indica- 

 tions. 



As the antenna of the newl> hatched larva has 

 1 8 joints aiul that of the adult male 78, 60 new joints 

 must be created during development. 



it should also be kept in mind that the first two 

 arhd the last i5 joints, corresponding together to the 

 whole anteiina except the 3rd joint, of th(> young 

 larva, })ersist to the advdt stage without any signs 

 of subdivision. It is thcrefoie in the portion between 

 joints r> and <vi that the nudtiplication must take 

 place. 



Besides thos(> represented in figs. ^1 and '), I 

 have been able to examine preparations of immature 

 larvae -m), o>, 3o and 3t mm. long respcM-tively. 



Male Larva, :>o ìniìt long. (fig. f)K — The an- 

 tennae are relatively very short, •?.-] mm. While the 

 length of the body is already one third of that of the 

 adult, the length of the antenna is only one seventh 

 of that of the imago. Moreover the first two joints 

 of the larva occupy one fifth of the total length 

 while the two basal joints of the imago onlv occupy 

 one eleventh. It is also clear that the antenna shbwn 



