49 



September 20th, between three and four hundred L. umbratus 

 workers were observed to hatch, some being killed as soon as 

 they emerged, and others after a day or two. None were allowed 

 by the L. niger workers to Hve longer than this, and some of the 

 pupse were stripped of their cocoons and thrown out of the nest. 

 Most of the dead workers were given to the larvae as food. The 

 following year the pupae began to hatch in the beginning of 

 July, and the young workers were not molested by their hosts. 

 By' the i8th there were twenty L. umbratus workers alive and 

 well, assisting the L. niger workers to tend the brood. None 

 were attacked or molested in any wav. unfortunately the nest 

 had to be left for over two months without attention during 

 Crawley's absence, and on his return the onh' ant alive was the 

 queen. A striking point was that there was not a single L. niger 

 amongst the hundreds of dead L. umbratus workers. 



The difficulty of obtaining newly-fertilised females of L. 

 umbratus prevented a repetition of this experiment until the 

 year 1908. In September 1908 Crawley found several deälated 

 females of L. umbratus wandering on a road near Nottingham. 

 Two were taken and enclosed in boxes with L. niger workers 

 from a queenless nest taken in July. The females, as in the 

 1896 experiment, instantly killed the workers. Another L. niger 

 was put with one of the females, who killed it also. Next day 

 another L. niger was put with her, and this time she was friendly, 

 so five others were introduced, and no hostility was observed — 

 indeed, the female and the workers saluted and caressed each 

 other. 



A few minutes after the introduction of the last worker, the 

 box was placed close to the door of the L. niger nest, and the 

 lid removed. The workers entered their nest, but the female 

 remained in the box. In less than two minutes after the tirst 

 worker entered the nest, more than sixt}' ants came out, found 

 the L. umbratus female, and surrounded her, amid great excite- 

 ment. Still surrounded by workers, she entered the nest, and 

 was evidently adopted as queen. 



The following morning the workers were found to have killed 

 all the young winged L. niger females in the nest, as in the former 

 experiment. The L. umbratus queen began to lay in May 1909, 

 and some of her eggs hatched in June, but only L. niger workers 



7 



