3 86 



THE QUARTERLY REVIEW OF BIOLOGY 



PRESENTATION OF DATA 



Brief inspection of the following data 

 taken from the records of 19x7 serves to 

 show how subtle are some of the factors 

 dealt with in the problem of controlled 

 mating of these insects. In all 96 virgin 

 queens were treated from once to ten 

 times each depending upon their behavior. 

 To prevent their leaving the hive at any 

 time in pursuit of natural mating the 

 hive entrances were always closed with 

 perforated zinc which allows workers to 

 pass in and out freely, but will not allow 

 queens and drones to pass. If a treated 

 queen was seen trying to leave the hive she 

 was taken to the laboratory and treated 

 again. Thus, some queens were fully 

 inseminated as the result of one injection, 

 and one queen is known to have received 

 only 50 per cent normal insemination 

 after nine injections. To ascertain at 

 once the extent of insemination the sper- 

 matheca was dissected out and crushed 

 under a microscopic cover slide. The 

 degree of insemination was then estimated 

 in per cent of the normal. 



Of the 96 queens that were treated 66 

 per cent received an appreciable degree of 

 insemination and 34 per cent received no 

 sperms into the spermatheca. Further- 

 more, of all the queens that received an 

 appreciable degree of insemination, 



79% received 5% normal insemination or more 

 38% received 10% normal insemination or more 

 17% received 50% normal insemination or more 

 11% received 75% normal insemination or more 

 6% received 100% normal insemination 



During the course of the season several 

 promising queens disappeared for one 

 cause or another before any brood was 

 produced, and there was no means of 

 determining whether they would have 

 been successes or not. In making the 

 records such missing queens were counted 

 as total failures although probably some 

 of them were inseminated. 



To estimate the degree of insemination 

 of queens that are still living, note was 

 taken of the percentage of worker-brood 

 produced by them. For example, if 10 

 per cent of her brood was sealed flat by 

 the workerbees, and the rest was sealed 

 oval, a queen was estimated to have re- 

 ceived about 10 per cent of the normal 

 degree of insemination. The absence of 

 any oval, or drone-brood, was taken to 

 mean that her insemination was normal. 

 This method of arriving at the results of 

 treatments is obviously open to criticism. 

 However, during the first few weeks of 

 her life as a laying queen, and especially 

 when a treated queen is being cultured in 

 a nucleus, there has been found to be a 

 remarkably close correlation between the 

 degree of her insemination and the propor- 

 tion of fertilized eggs that she lays. 



WHY ARE RESULTS SO VARIABLE? 



That instrumental insemination of queen- 

 bees is still in the experimental stage is 

 evidenced by the large proportion of 

 queens that receive only partial insemina- 

 tion as the result of a single injection. 

 It still frequently happens that when half 

 a dozen queens from the same hatch are 

 treated exactly alike on the same day, 

 some will be copiously inseminated while 

 others are complete failures. Repeated 

 injections build up the degree of insemina- 

 tion, but repeated injections are imprac- 

 tical in most hereditary studies. Queens 

 that have received less than 10 per cent 

 normal insemination cannot be depended 

 upon to settle themselves to oviposition; 

 however, as low as 5 per cent of worker- 

 brood may be quite sufficient to enable 

 valuable genetical studies to be made. 

 To the question, why are the results so 

 variable, we are compelled in candor to 

 answer that we do not yet know. 



In systematic search for an underlying 

 cause of such wide differences in results, 

 both the technique and the general condi- 



