108 Papers from the Department of Marine Biology. 
texture would lead one to suppose. Unless the arms were voluntarily 
ruptured, they were not easy to break, and even pinnules and cirri 
withstood hard pulling, when specimens were being untangled from 
each other. The arms were seldom voluntarily ruptured until the 
specimens had been in the laboratory for some hours, and if properly 
supplied with fresh, clean water, individuals could be kept for a day 
or more. The diversity of color, the delicacy of structure, and the 
gracefulness of movement surpass description and easily justify Mr. 
A. H. Clark’s statement (1908, Geog. Jour., London, p. 606) that among 
marine invertebrates they are unrivaled. They may well be called 
‘‘sea-lilies,”’ for nothing outside the most beautiful products of the 
vegetable kingdom can be compared to them for beauty. 
METHODS. 
It proved impracticable to have aquaria at the laboratory large 
enough to keep comatulids in a healthy condition, but a compartment 
in the live-car anchored just outside the reef served as a home for 
numerous specimens, available for experiment either at the car or at 
the laboratory. Crinoids kept in the car for many days lost the 
terminal portions of some or all of their arms, apparently because of 
the constant swaying of the car, and during the last week of our stay 
a particularly rough night killed nearly all the comatulids in the car. 
For observations and experiments at the laboratory, a tank or trough 
34 feet long, 10 inches wide, and 10 inches deep, made of metal and 
wood, answered very well. This will hereafter be referred to simply 
as the tank. 
As only a part of the time could be given to the work, even for the 
short period we were at Maér (five weeks), no attempt was made to 
secure quantitative results. Consequently only superficial attention 
was given to variations in the strength of stimuli or in rapidity of 
response. The purpose was simply to learn how comatulids reacted 
under normal conditions, so that such observations as were made on 
responses to chemical stimuli were purely incidental. Of course, obser- 
vations were continually made on the reef and some experiments were 
undertaken there, but various circumstances combined to make the 
results negative. Had more time been available, even another week, 
I would have been able to profit by the experience of some of these 
failures, but as it was they were of very little significance, although 
they will be referred to further under the discussion of locomotion. 
Like other students of animal behavior, I found individual differences 
sufficiently marked so that selected specimens gave the best results. 
After a number of specimens had been observed and frequently dis- 
turbed, one or more were picked out, which seemed most responsive, 
and the more careful observations were made on them. There is no 
reason for believing such observations unrepresentative, for these 
