OBSERVATIONS ON THE PROGENY OF VIRGIN ANTS. 357 



The absence of indubitable proof that the unfecundated ant- 

 egg produces either male or female ant impelled me to undertake 

 the formation of ant-groups in which no member had ever lived 

 with a male, and from whose abode males were excluded. In 

 the summer of 1904 I sequestered pupae, \v'th two or three 

 workers to take care of them, in artificial nests in which there 

 were neither queens nor males, and as soon as these pupae 

 hatched I segregated the callows in new nests into which no egg, 

 larva or pupa was ever introduced. My ant-groups were thus 

 made up of workers indubitably virgin, and the eggs deposited in 

 their nests were certainly unimpregnated. 



GROUP A. Camponotus pictus. 



Group A consisted of thirty workers, majors and minors, of 

 Camponotus liercnlcamts pictus, hatched between July 11 and 31, 



1904, and kept in segregation from their hatching until October, 



1905. Their first eggs, ten in number, were deposited between 

 May 14 and 18, 1905, and these had increased to about fifty on 

 June 4. The first larva appeared on June 6, nineteen days after 

 the first eggs were observed. On July 7 the larvae had made 

 notable progress in number and in size, the largest then being as 

 long as an adult worker. The first cocoon was spun on July 16, 

 and the first offspring of these segregated virgin workers appeared 

 on August 14. Between August 14 and September 30 their 

 cocoons gave forth thirty-two notably large and sturdy males. 

 That no female might escape observation if hatched from these 

 cocoons, the cocoons were transferred, soon after their formation, 

 to an annex of the nest where only five workers were admitted. 

 Cocoons and nurses were daily counted, and it is certain no 

 queen nor worker ever hatched from these cocoons. 



GROUP B. Formica argcntata. 



Group B consisted of about fifty Formica argentata workers, 

 all hatched from sequestered cocoons during September, 1904, 

 and kept in segregation by me from the time of hatching until 

 after the close of this series of observations. From eggs de- 

 posited on July 7, 1905, the first larva appeared on July 19. 

 The first cocoon was spun on August 4, and the first ant hatched 



