440 PROCEEDINGS OF THE ACADEMY OF [July, 



dorsal side of the abdomen became grayish, and on the twelfth 

 day the head, thorax and abdomen were slate color, while the 

 limbs remained white. Thereafter the color deepened to black, 

 and in each of the twelve cases the pupa became an ant on the 

 seventeenth day. 



Four of these dwarf males, that remained during their natural 

 lives with their worker-progenitors, lived respectively fifteen, thirty- 

 two, thirty-four and forty-five days. 



The food which the Stenamina fulvum j^iceum were seen to eat in 

 captivity was, in the order of their apparent preference, fragments 

 of flies, roaches, mealworms, beetles and spiders ; morsels of 

 sponge-cake, white bread moistened with sweetened water or white 

 syrup; apple, banana, boiled sweet-potato ; fat of boiled fresh 

 beef; soft gum-drop, almond paste, pie crust, hickoiy nut and 

 honey. They showed no lively interest in other than insect food 

 of which they had been for some days deprived. They appeared 

 to avoid all raw or cooked meats other than particles of fat. 

 Their liking for a varied diet and their attention to unusual deli- 

 cacies indicate a highly developed sense of taste. 



Though the attitude ordinarily assumed in eating is that of 

 standing on all six legs and lapping the food, I have twice seen 

 an ant stand on four legs, using the front feet to hold an insect- 

 egg to its mouth, suggesting the posture in which a squirrel com- 

 monly eats nuts. 



The amount of food required to sustain life must be small. I 

 isolated sixteen workers in groups in clean Petri cells, containing 

 nothing but sponges that were frequently cleaned with ninety-five 

 per cent, alcohol and then saturated with water. Of these ants one 

 lived five days, five lived six days, two lived seven days, one lived 

 eight days, three lived nine days, one lived twelve days, one sixteen 

 days, one twenty-one days, and one thirty-four days without visible 

 food. That these ants died from starvation and not from other 

 cause was indicated by the conti'ol experiments in which other ants 

 similarly placed, but with a supply of food, continued to live on for 

 months. The ability of Stenamma fulvum to endure starvation 

 is, however, exceeded by the less active Formica fusca and the 

 sluggish Ponera coardata, one of the former haviag lived in my 

 Petri cell forty-one days, and one of the latter forty-three days, 

 ■without visible food. Formica sanguinea shows lesser tenacity of 



