994 THE AMERICAN NATURALIST. [Vor. XXXV. 
the altered conditions and so acquire a more or less completely 
terrestrial habit. At any rate this thigmotactic reaction to 
the surface tension of water must be regarded as one of the 
causes which may induce creeping aquatic forms to leave the 
water. 
CHEMOTAXIS 
Among the chemotactic reactions the most conspicuous 
under conditions approaching the normal are the reactions to 
the composition of the water. In general the animals show 
some tendency to aggregate where the oxygen content is 
greatest. Placed in jars with masses of alge and other ani- 
mals, and kept where formation of oxygen by the plants does 
not occur in any marked degree, the animals aggregate at the 
surface as the water becomes foul Under these conditions 
they are found either among the algze just beneath the surface 
or crawling or resting beneath the surface film or about the 
margins. Furthermore, as the water becomes foul, animals 
which did not leave it earlier now crawl up the sides of the 
glass and die there. This reaction is in this case doubtless a 
mixed negative chemotactic and positive thigmotactic reaction ; 
that is, the impulse to seek the surface is given by the increas- 
ing lack of oxygen or the increasing quantity of decomposition 
products in the water. Having reached the surface and being 
perhaps still restless in consequence of the chemical stimulus 
from the composition of the water, the worm leaves the water 
in response to the combined stimuli, and thus dies more quickly 
than it would if it remained in the water, since it possesses à 
high degree of resistance to the effects of foul water. 
ENEMIES. 
In general, Stichostemma appears not to suffer greatly from 
the attacks of predatory animals. Mutilated specimens are 
rarely found, though experiments show that even relatively 
small pieces of the body may live for weeks. Montgomery 
(95) stated that cyprids destroy Stichostemma eilhardi. Under 
ordinary conditions I have never been able to observe that 
