256 PARTHENOGENESIS VII 



and some of those which promised the best results and 

 had cost the most pains came to a bad end in the 

 very last days of the research. In order to determine 

 the sex of the wasps born from the eggs laid by the 

 parthenogenetic females, Siebold in most cases only 

 allowed the develo23ment to proceed sufficiently far to 

 enable him to recognise the sex by anatomical investi- 

 gation. The dried skin, however, of such grubs as 

 were found dead in their cells afforded sufficient evi- 

 dence of the sex. In all cases the parthenogenetic 

 offspring was without exception male. The queen- 

 wasps, as Ave have mentioned, also late in the season 

 lay eggs which produce drones, which are easily dis- 

 tinguished from the drones parthenogenetically pro- 

 duced by their larger size. It occurred to Siebold 

 when he first ascertained that the queens produce 

 drones that such drones might visit his virgin colonies, 

 and thus his whole experiment be invalidated. He 

 was, however, reassured on this point by a nearer 

 acquaintance with the Polistes ; for such drones are 

 not born till later than the period at which his small 

 females laid their eggs, the former event taking place 

 at the end of July, the latter at the beginning ; and, 

 furthermore, as w^e have noticed above, it is not till 

 still later (August), when the experimental cells were 

 long since all occupied with eggs, that the power and 

 desire of sexual activity comes to these drones. 



Siebold's experiments extended over four years, 

 and although some hundreds of nests were more or 

 less observed, only thirty -seven — but these amply 

 sufficient — gave the answer to his questions, passing 



