THE WEB OF LIFE 331 



do not look after the larvae ! They require some gustatory 

 reward for their apparent altruism. 



It is instructive to notice M. Pieron's general conclusion 

 that, as ant-evolution becomes more complex, the members 

 of the community become more and more dependent on 

 one another. The species which are most thoroughly 

 self-sufficient are the most successful species, as far as 

 numbers and distribution are indicative of success. On 

 the other hand, those that show slave-keeping and parasitic 

 habits have smaller numbers and sparser distribution. 

 ' It is evident that when ultra-civilization degenerates 

 into slavery and parasitism it is neither good for man nor 

 ant '. 



The records of studies on ants make quite a good-sized 

 library, but we have looked into the ant-hill enough for 

 our purpose of iUustratiag some of the features of an animal 

 society — division of labour, subordination of the individual 

 to the whole, a capacity for unified action and co-operation. 

 As we have had to say so often, the more we know about 

 it the more the wonder grows. The social life seems so 

 intricate that we wonder how it could have evolved at all. 

 Yet let us look at it in one of the simplest expressions. 

 There is a Mediterranean ant, Aphwnogaster sardoa, which 

 illustrates what may be called an incipient societary form. 

 According to Dr. Krause-Heldrungen these ants live in 

 holes in the ground and do not build. Nor do they store 

 or entertain guests. Huddling together is their form of 

 sociahty. They form living balls, ant interlocked with 

 ant by the mandibles and tarsal joints, and they hold the 

 eggs, larvae, and pupae in the middle. It is almost like 

 a diagram of a primitive society and certainly matri- 

 archal ! A ball consists of three hundred to a thousand 



