THE ANTS OR PISMIRES 



parallel with the simplest kind of natural polymorphism, can 

 be produced experimentally. 



In the past, the chief support for the trophic theories has 

 been derived first from what is supposed to occur in bees and 

 wasps ; in these caste-production has been supposed to be 

 entirely due to differential feeding. Secondly, in most ants 

 the polymorphism of the workers is mainly a matter of size. 

 Species with a sharply discontinuous soldier caste are few. 

 Thus the most usual type of variation within the worker 

 caste is very much what would be expected from an unequal 

 division of food. It is known from observation nests that 

 worker ants may give exceptional attention to grubs which 

 happen to be larger than the others, so that a slight natural 

 variation in size might be accentuated by the behaviour of 

 the nurses. 



Recently, some experimental evidence has been obtained 

 which supports the trophic theory of the origin of castes. 

 Goetsch and Gregg have studied the relation of food supply 

 to the production of soldiers in Pheidole. This is an ant in 

 which the large, swollen-headed soldiers are sharply differen- 

 tiated from the small workers; no intermediates between them 

 are found. Goetsch showed that a group of grubs given 

 honey or some other liquid food turned into workers, where- 

 as a group provided with pieces of dead insects produced 

 one or more soldiers ; the number depended on the quantity 

 of solid food. Liquid food is taken into the worker crop and 

 then some is regurgitated for each grub individually, so that 

 each one gets a more or less similar share. Solid food, on the 

 other hand, is merely put near the grubs, so that one or more 

 of them, by reaching it and feeding themselves, get a much 

 bigger share than the others. This solid food must be received 

 during the first five days of life, otherwise the grub irre- 

 vocably becomes a worker. The feeding on the fourth and 



141 



