354 Experimental 'Zoology 



incompatible with Dzierzon's theory; yet several explanations 

 have been offered to account for these exceptional cases, for 

 it is generally admitted that they are exceptional and do not 

 represent what usually takes place. 



In the first place, the pedigree of the queen may not have been 

 above suspicion, and consequently some of her offspring may 

 show traces of their mixed origin. It is said that at the present 

 time the domesticated races of bees have been so often crossed 

 that it is difficult to obtain a pure individual. In any particular 

 instance the matter could be easily tested by rearing sister queens 

 of the same brood, crossing one and breeding the other true. 

 This simple test does not appear to have been made. 



In the second place, the hybrid drones may have come from 

 the hybrid workers that sometimes lay eggs, rarely in the com- 

 mon hive bees, but more often in the Syrian and Palestine races. 

 In the third place, it has been found that Italian bees of presum- 

 ably pure stock often produce drones that are very variable in 

 their markings. In a recent paper by Casteel and Phillips it 

 has been pointed out that the males are often more variable than 

 the workers a fact of much biological interest in itself, aside 

 from the present question, since it shows that variation is not 

 necessarily connected with the union of two germ-cells from 

 different individuals, as Weismann and his school have assumed. 



In another way Dzierzon's theory was tested, and apparently 

 found wanting. Landois (1867) put worker eggs into drone 

 cells, and drones were produced. Conversely he put drone eggs 

 into worker cells and workers were produced. The real expla- 

 nation of this result is, however, not that the kind of cell de- 

 termines the sex of the individual, as the experiment seemed to 

 show, but that the bees themselves often clean out a cell into 

 which an egg has been introduced, perhaps because of the 

 imperfect method of fixation of the egg, and subsequently the 

 queen will deposit the proper kind of egg in the cell. Thus 

 the bees themselves easily defeated the purpose of the experi- 

 ment. The experiment could be properly carried out by 

 removing the queen when the transfer is made, so that no 



