116 Papers from the Department of Marine Biology. 
crannies in the rock, but generally the movement was over the surface 
of the rock to its margin and thence down into the shade of its now 
lower side. The movement was never spasmodic or hurried, but unless 
interfered with was very steady and soon accomplished its purpose. 
It is difficult to see how this invariable habit can be interpreted as 
anything else than a response to light as a stimulus. Since the rock 
might be 3 feet under water, no change of temperature could possibly 
have caused the movement; it is conceivable, however, that the change 
of position due to the overturning of the rock was the effective stimulus. 
But the theory that change of position is the determining factor seems 
to be eliminated by the facts that among the living corals comatulids 
occur with the oral surface either up, down, or lateral, and that when 
the coral around them is broken away with a hammer, exposing them 
to strong light but not disturbing their position, their movements are 
also away from the lighted area into the shadow. 
Experiments in the laboratory were not altogether satisfactory, so 
much individual difference was shown. As a rule, if a comatulid was 
placed at the middle of the tank, one end of which was covered over, 
while the other end was in sunlight, movement to the shaded end 
almost always was the first result. Often, however, the animal did 
not remain there, but moved about restlessly, even to the brightly 
illuminated end. In such cases it is a fair inference that the absence 
of suitable surface for attachment was the cause of the restlessness. 
More consistent results followed when direct sunlight was allowed to 
fall on comatulids which had been in the shade. With scarcely an 
exception, they moved out of the brightly lighted area. Experiments 
with artificial ight were entirely negative, presumably because of the 
weakness of the stimulus, nothing better than an ordinary kerosene 
lamp being available. No specific differences were noted with reference 
to light reactions, all the species watched giving the same results. 
Bright sunlight thrown abruptly on a comatulid lying in shadow often 
produced immediate and sometimes more or less spasmodic closing of 
the arms over the oral surface of the disk, as though that part of the 
animal were most sensitive to the stimulus. In confirmation of the 
view that the disk is more sensitive to light than the arms is the fact 
that in their natural habitat the crinoids often extended the arms out 
into the light, while the body itself was in heavy shadow. Indeed, all 
along the southwestern reef, on the brightest days, they were to be 
seen in such a position. 
RESPONSE TO HEAT. 
As ice was not obtainable at Maér and it was difficult to lower the 
temperature of any large amount of water to any considerable degree, 
my only experiments with regard to the response of comatulids to 
temperature changes dealt with the effect of raised temperatures or, 
