THE PLASTIC BEHAl'lOR OF ANTS. 543 



the food, constantly climbing up and down the legs, but only a few 

 attempted to cross the oiled bandages and these were not successful. 

 The following morning the table was well covered with ants. They 

 had gone up the wall over the first trail and passed on up to the ceiling, 

 then over that diagonally until they were over the table, when they 

 dropped down onto it. Very few ants were noticed returning from 

 the ceiling, but a constant stream of them was going up. At the point 

 where the table had formerly touched the wall quite a number of ants 

 were clustered, evidently at a loss to know where to go. The ants, on 

 leaving the table, usually went down one of the legs and were crossing 

 the coal-oil bandages with apparently little or no injury to themselves. 

 Some dropped directly from the table top to the floor." The facts 

 related by Titus are not to be questioned, but the same criticism applies 

 to this case as to the preceding: the whole series of events was not 

 observed. Moreover, although ants are often seen to drop from plants 

 or walls, either voluntarily or when disturbed, we know nothing as yet 

 about the various instinctive adaptations which such behavior may 

 involve. It may be a much more frequent method among ants of clear- 

 ing vertical distances than has been supposed. Hence, I believe that 

 instead of attributing such acts to reasoning, it would be wiser to sus- 

 pend judgment till careful experimental data are available. 



In conclusion it may be noted that all the activities of ants, their 

 reflexes and instincts, as well as their plastic behavior, gain in precision 

 with repetition. In other words, all their activities may be secondarily 

 mechanized to form habits, in the restricted sense of the word. This 

 is tantamount to saying that even the reflexes and instincts are not so 

 stereotyped but that they may become more so by exercise during 

 the lifetime of the individual. And not only do ants thus form habits, 

 but, as several myrmecologists have observed, these habits when once 

 formed are often hard to break. It is certain that many instincts 

 among the higher animals are at first incomplete or indefinite and are 

 guided into their proper course by stimuli that affect the organism at a 

 later period. This is probably true also of many formicine instincts. 

 There is little doubt, moreover, that the more fixed or stereotyped 

 instincts are phylogenetically the older. This fact, and the close super- 

 ficial resemblance of habits to instincts, has led many authors to derive 

 the latter from the former. The views on the origin of automatic 

 behavior, however, are so diverse and conflicting that they cannot be 

 satisfactorily considered without entering into a discussion of the doc- 

 trines of the Neodarwinians, Neolamarkians and those who believe in 

 coincident, or organic selection. In my opinion we have little to gain 

 at the present time from such a discussion. As Bergson says : " It is 



