610 PROCEEDINGS OF THE ACADEMY OF [Sept.^ 



that colonies apparently alike, and with nest-exits but a few feet 

 apart, sometimes show bitterest hostility to one another. It also 

 explains the fact that ants taken from the extremes of an ant-domain 

 fifty yards in diameter, and kept apart for two years, amicably 

 reunite, provided that no young has hatched in either segregated 

 group. It furnishes also explanation of the highly variable 

 behavior of the ants of the same lineage toward one another when 

 brought together after segregation ; and it enables one knowing the 

 exact conditions to accurately prophesy wdiat a given ant will do 

 on meeting another given ant of her own lineage. 



In the following experiments, which are but a few examples from 

 among many made by me in the summer of 1902 for the purpose 

 of ascertaining the cause of hostility between colonies of the same 

 species, all the ants were of the C colony, much used by me 

 because of easy access to its populous natural nest. 



For the ants a year or more old, I used queens and workers of 

 this colony that were captured by me on August 22, 1901, and 

 that have since been kept by me in small artificial nests, where no 

 pupa has been allowed to hatch, and to which no ant has ever been 

 introduced from outside. 



When the ants did not differ in size or color sufficiently for me 

 to thereby distinguish them, I marked those that I desired to 

 observe.^ 



The ants depended upon to show the feeling of their kind toward 

 such as were introduced, were all engaged in the care of inert 

 young in Petri cells, where they had been for some days or weeks 

 established as a family group. When I was about to use any such 

 cell in experiment, I sometimes introduced an ant of an alien 

 colony in order to ascertain the presence of fighters among the 

 residents of the cell. Such aliens were always removed before the 

 experiment. 



The ants could not carry intruders outside the cell, unless I lifted 

 its cover. I therefore considered, in these experiments, the seizing 

 and dragging of any ant away from the larvse pile and across the 



s I mark the ants upon the top of the abdomen using dried varnish, 

 into which water-color paint has been rubbed before drymg. The var- 

 nish is softened for immediate use by a drop of chloroform. The ant is 

 "•ently held by the head until the chloroform has evaporated from the 

 dab upon her abdomen. This mark often remains upon an ant several 

 -weeks, and does not appear to influence her action nor that of her 

 associates in the experiment. 



