48 



more likely L. umhratus of the fiist brood, which always consists 

 of small ants. 



Adlerz's explanation of this mixed colony was that the L. 

 niger workers had pillaged some L. flavus pupse and hatched 

 them. This is, however, very improbable. 



Craw^ley first proved by experiment that queenless colonies 

 of L. niger will accept fertile females of L. umbratus, and bring 

 up the ofíspring of the latter until the colony becomes a mixed 

 one of the yellow and black ants. 



In August 1896 he procured a nest of L. niger containing a 

 queen, near Oxford. There were about 400 workers, worker 

 and female pupœ, and a large quantity of eggs. The queen, 

 however, had been injured during the operation of digging up 

 the nest, and died. A fortnight later he picked up a newly- 

 deälated female of L. mnbratus in the same neighbourhood, 

 and placed her in a box with two L. niger workers, which she 

 immediately killed. A day or two later, however, she was 

 perfectly friendly with some other L. niger workers. She was 

 thereupon put in a box with four workers from the captive and 

 now queenless colony, and was also friendly to them. Later 

 on the same day he removed the lid of the box and placed it 

 near the door of the nest. The workers immediately entered, 

 and were followed in a few moments by the queen, who entered 

 the nest of her own accord. In a few seconds the whole nest 

 was in a turmoil, swarms of ants collecting round the female 

 and saluting her. One ant only attacked her for a short time, 

 and she was soon completely hidden under a mass of workers. 



A few days later the ants killed the young winged L. niger 

 females that had come to maturity in the nest. L. umbratus 

 females do not appear to lay eggs until the 3'ear following their 

 impregnation, and this queen began to lay on June 26th, 1897, 

 so that all the ants that came to maturity in this year were 

 L. niger. The next year also all the workers produced were 

 L. niger, from parthenogenetic eggs laid by workers. The offspring 

 of the L. umbratus queen must have been devoured in the egg 

 or larval stage. 



In 1899 the eggs of the L. umbratus queen were at last allowed 

 to reach the pupal state, and the fragments of an L. umbratus 

 callow were discovered on August loth. From this date up to 



