76 A NATURALIST IN HIMALAYA 



downwards over its back under the influence of gravity 

 until its position exactly resembles that of a Cremasto- 

 gaster ant. A large abdomen, under such insufficient 

 muscular control as to have its position affected by the 

 stress of gravity during the ascent and descent of the 

 tree-trunk, must to some extent diminish the stability 

 of the insect. And I suspect that a Cremastogaster 

 ant, which is continually moving up and down the 

 trunks of the trees, has found it more advantageous 

 to its stability and balance to maintain its abdomen 

 permanently in the position that gravity would fix it 

 during the descent of a tree, rather than drag behind 

 it an unwieldy mass ready to sway about in every 

 direction that might always serve as a threat to its 

 equilibrium. The adoption of this attitude might be 

 expected to produce anatomical changes in the struc- 

 ture of the ant. This seems to be the case, for the 

 narrow pedicle, that in most ants connects the thorax 

 to the middle of the front of the abdomen, is in this 

 genus attached not to the middle but to the upper 

 margin of the abdomen. An attachment of this nature 

 permits the unusual attitude to be more easily adopted. 

 The migrations of ants often display the plasticity 

 of instinct in the species and the power of the ants to 

 modify their behaviour to meet unusual conditions. 

 One small and agile little ant, Acantholepis frauenfeldi, 

 seemed to be continually in a state of migration (see 

 Plate, p. 60). After every shower of rain a stream of 

 workers might be seen hurrying from an old to a new 

 nest, all heavily laden with larvae. I suppose that the 

 rain, flooding the nest, makes it uninhabitable for the 

 ants, and they are forced to remove to drier quarters. 

 The ants often do not migrate to any great distance ; 



