THE INSTINCTH'E BEHAl'IOR OF ANTS. 521 



laboratories. Spencer, Loeb, Bethe, Driesch and others have supported 

 this hypothesis with much clean-cut argument ; but the way to its con- 

 ception was prepared long ago by Claude Bernard in his description 

 of certain complex physiological activities, like those of the alimentary 

 tract, as " mouvements reflexes regulierement enchaines." The diffi- 

 culty with the hypothesis is its schematism, for while we may admit 

 that an instinct may be described as a compound reflex, we must also 

 admit that it is more than this, since the reflexes are not merely strung 

 along in sequence, but constitute an organized system of coordinated 

 activities which coimplicate and interpenetrate one another, so to speak, 

 and grow and change by modification in toto, or by intersusception and 

 not by simple apposition of new activities. 



Biologists find it increasingly difficult to draw a hard and fast line 

 between instinct and reflexes, or between either of these and the simple 

 vital activities of protoplasm. The definition of instinct cited above 

 is perfectly applicable to a unicellular organism, or to a single Metazoan 

 cell, considered as a whole. It is difficult or impossible, moreover, as 

 Loeb and Driesch have insisted, to dissociate the instinctive activities 

 from those of growth and development. This is due to the fact that 

 instinct is so intimately and inextricably involved in the structure of 

 the organism. As Bergson says: "It has often been remarked that 

 most instincts are the prolongation, or better, the achievement, of the 

 work of organization itself. Where does the activity of instinct begin? 

 Where does that of nature end? It is impossible to say. In the meta- 

 morphoses of the larva into the nymph and into the perfect insect, 

 metamorphoses which often require appropriate adaptations and a kind 

 of initiative on the part of the larva, there is no sharp line of demarca- 

 tion between the instinct of the animal and the organizing work of 

 the living matter. It is immaterial whether we say that instinct organ- 

 izes the instruments which it is going to use, or that the organization 

 prolongs itself into the instinct by which it is to be used." The spin- 

 ning of the cocoon by the larval ant is a good example of the kind of 

 instinct to which Bergson refers. From one point of view this is 

 merely an act of development, and the cocoon, or result of the secretive 

 activity of the sericteries and of the spinning movements of the larva, 

 is a protective envelope. But an envelope with the same protective 

 function may be produced by other insect larvae simply as a thick, chiti- 

 nous secretion from the whole outer surface of the hypodermis. Here, 

 too, we have an activity which, though manifested in a very different 

 way, is even more clearly one of growth and development. And when 

 the workers of (Ecophylla or Polyrhachis use their larvae for weaving 

 the silken envelope of the nest, as described in Chapter XIII, we have 



