DONISTHORPEA. 197 



Therefore, from the above facts alone, it seems doubtful whether 

 the female of this species can found a colony unaided. 



But further observations in the field point to Donisthorpea umbrata 

 and D. mixta as the host-species of D. fuliginosa. In 1897 I found 

 at Lymington a large colony of D. fuliginosa in a hollow tree, 

 and D. umbrata was undoubtedly living with it, as workers of 

 both species were going in and out of the same holes 35 . Crawley in 

 1898 repeatedly found workers of D. umbrata walking unmolested 

 with the workers of a large nest of D. fuliginosa established under 

 his house near Oxford 53 . In September, 1900, Tuck sent to me a 

 worker of D. umbrata taken in a nest of D. fuliginosa at Bury 

 St. Edmunds 54 . 



In 1904 de Lannoy 46 found at Knocke-sur-Mer a few workers of 

 D. mixta in the midst of a large colony of D. fuliginosa which 

 were on good terms with the workers of the latter, and in 1906 he 

 again found workers of D. mixta in several D. fuliginosa nests. 

 Forel 47 , Emery 48 , Wasmann 50 , Wheeler 51 , and I 54 commented on 

 de Lannoy's observations, and expressed the opinion that the 

 presence of these D. mixta workers was due to the fact that fertile 

 D. fuliginosa females had entered nests of the former species and 

 been accepted. The queens of the D. mixta had then died, or had 

 been killed, either by their own workers, or by the D. fuliginosa 

 females, and the offspring of the latter were reared by the D. mixta 

 workers. In the course of time many of the latter had died off, 

 and the few found in the nests were the survivors of the original 

 D. mixta colonies. 



Crawley and I determined to test this hypothesis by experiments 

 on captive colonies. In July, 1910, a portion of a nest of D. fuli- 

 ginosa was dug up at Darenth Wood, containing a quantity of 

 workers, larvae, males, and winged females, but no queen. The 

 ants and brood were divided into two equal portions and each 

 established in a four-chambered " Janet " nest. (It may be here 

 mentioned that this ant cannot be kept in close confinement, but 

 that if an observation nest be connected by long glass or india- 

 rubber tubes to another plaster nest, or glass bowl, or some other 

 contrivance in which their food is placed, the ants will thus obtain 

 sufficient exercise.) During July all the males died and most of 

 the females, with the exception of about twelve, which were found 

 to be dealated. As some of these latter subsequently laid eggs, 

 from which larvae were reared, it is highly probable that mating 

 had taken place inside the nests. 



In the beginning of December, 1910, a nest of D. umbrata without 

 a queen was obtained at Weybridge and divided into two equal 

 portions, which were established in Janet nests. 



The first experiment was made on December 10th when one of 

 the dealated female fuliginosa was placed in the light chamber of 

 one of the umbrata nests. She immediately entered the most 



