812 THE AMERICAN NATURALIST. [Vor. XXXV. 
instincts appear as modes of that simplest of all forms of 
activity, the reaction of a living cell to the stimulus received 
from its environment. And we are led to conclude with 
Loeb ('00, p. 177) that “the discrimination between reflex and 
instinctive actions is chiefly conventional.” The work of this 
last author, above all others, is of great value in freeing us 
from some of the traditional misconceptions of instinct. He 
has succeeded in tracing a number of instincts to simple trop- 
isms (or taxes) and has shown good reasons for maintaining 
that many of the more complicated instincts are only catenary 
reflexes (Kettenreflexe). These, however, have not yet been 
sufficiently analyzed. 
Loeb’s conception is also fruitful in another direction, for 
as Mach says (00, p. 64) it throws light on the relations 
between development and instinct. Numerous “ instincts " of 
ants, such as their reactions to moisture, heat, light, and con. 
tact, are evidently simple reflexes and may, I believe, be 
treated as cases of hygro-, thermo-, helio-, and thigmotaxis. 
The first and last of these reactions are especially striking. 
Similarly the olfactory reactions, which are such an extremely 
important factor in the lives of ants, are probably not essentially 
different from the chemotactic reactions of simpler organisms. 
The consociation of ants in mixed and compound nests is 
undoubtedly dependent to a very considerable extent on 
olfactory reactions. The young auxiliaries that hatch from 
stolen pupz are at once adopted in the mixed nests because 
they have acquired the nest odor of the dominant species. 
Moreover, the fact that these intimate relations are established 
only between ants of rather close taxonomic affinities is prob- 
ably due to their having very similar odoriferous secretions to 
begin with. The animosity of ants seems to be at once excited 
by species which emit peculiar or unfamiliar odors. While I 
accept Loeb’s conception as simplifying to a considerable 
extent the problem of instinct, I cannot suppose with Bethe 
('98,'00) that the behavior of ants and bees is entirely of , 
reflex nature; I can only indorse Wasmann's (99b) and Forel's 
('00—01) comments on this author's extreme views. : 
Wasmann (91, p. 179 e£ seg.) has shown in detail why ts 
