120 The Study of Animal Life part i 



was very deliberate in the steps which led to the first 

 domestications. (2) Nor is it likely that the process 

 began in a casual way, and that it became predominant 

 in four or five species in the course of natural selection.. 

 For the habit is more a luxury than a necessity, and 

 it is not likely to have been evolved before the estab- 

 lishment of the sterile caste of workers, wjtio h ave no me ans 

 of transmitting their experience. Moreover, initial steps are 

 always difficult to explain on this theory. (3) The theory 

 which seems to me warrantable is that the habit arose by 

 a gradual extension of habits previously established, that it 

 was neither deliberate nor casual in its origin, but a natural 

 growth, beginning neither in a clever experiment nor in a 

 fortunate mistake of an individual worker ant, but the 

 outcome of the community's progressive development in 

 "intellectual somnambulism," helped in some measure by 

 the sluggish habits of the aphides. And, if you wish, the 

 formula may be added, " which was justified in the course 

 of natural selection." 



3. Storing. — Not a few animals hide their prey or their 

 gatherings, and with marvellous memory for localities 

 return to them after a short time. But genuine storing 

 for a more distant future is illustrated by the squirrels, which 

 hide their treasures like misers. Many mice and other rodents 

 do likewise, and in some cases the habit seems to become 

 a sort of craze, so large are the supplies laid in against the 

 winter's scarcity. Very quaint are the sacred scarabees 

 (Ateucus sacer), which roll balls of dung to their holes, and 

 sometimes collect supplies at which they gnaw for a couple 

 of weeks. Some ants (e.g. Atta barbara) accumulate stores 

 of grain, occasionally large enough to be worth robbing ; 

 and there is no doubt that they are able to keep the seeds 

 from germinating for a considerable time, while they stop 

 the germination after it has begun by gnawing off plumule 

 and radicle and drying the seeds afresh. Dr. M'Cook's 

 account of the agricultural ant of Texas (Pogomyrtnex 

 barbatus) gives even more marvellous illustrations of 

 farming habits, for these ants to a certain extent at least 

 cultivate in front of their nests a kind of grass with a rice- 



