840 
MR. G. J. ROMANES AND PROFESSOR J. C. EWART 
peculiar, and they may best be studied by taking the animal out of the water, placing 
it upon a dry flat surface, and watching the movements of its feet by placing the eye 
on a level with them. It may then be observed that the mode of locomotion is as 
follows:—The animal points all the feet of all the rays in the direction of advance, and 
then simultaneously distends them with fluid; they thus become so many pillars of 
support, which raise the animal as high above the flat surface as their own perpen¬ 
dicular length. The fluid is then suddenly withdrawn, and the Star-fish falls forward 
flat with a jerk. This manoeuvre being again and again repeated at intervals of about 
a quarter of a minute, the animal progresses in a uniform direction at the rate of about 
an inch per minute. It is particularly noteworthy that, in this mode of progression, 
all the feet of all the rays are co-ordinated in their action for determining one definite 
direction of advance—those in the ray facing that direction acting forwards, or centri- 
fugally, those in the hinder rays backwards, or centripetally, and those in the lateral 
rays sideways. 
When the animal is walking along a flat horizontal surface in water, its mode of 
progression appears to be the same as it is on a dry surface, only the motion of the 
feet is now so rapid that there is a considerable difficulty in following it with the eye. 
It appears, however, as if the feet, besides being used as walking-poles in the manner 
just described, are also used to sweep backwards along the floor of the tank, and so to 
assist in propelling the animal forwards after the manner of cilia. Therefore, while 
walking in water, this Star-fish is kept stilt-high above the surface on which it is 
walking, by some of its feet, while others of its feet are engaged in these sweeping- 
movements. 
The result of all these movements is to produce a kind of locomotion which would 
seem more suited to a Centipede than to a Star-fish, and the suggestion that it is so is 
borne in upon the mind all the more forcibly by the surprising rate at which the 
animal is thus enabled to move. For while an ordinary Star-fish only crawls at the 
rate of 2 or 3 inches per minute, Astropecten can crawl, or perhaps more correctly run, 
at the rate of between 1 and 2 feet per minute. 
When placed upon their backs, the righting movement of these Star-fish are 
performed by raising the disc from the floor of the tank, till the animal rests only on 
the tips of its five rays (Plate 81, fig. 21). Two rays—for instance, 4 and 5—are then 
bent under the disc, while 2 and 3 are raised on a level with the disc. The disc 
becoming tilted in the direction of 4 and 5, 2 and 3 are now thrown over the disc, 
and assist by their weight in revolving the whole system upon an axis situated at 
about the level A, A. This mode of executing the righting manoeuvre is somewhat 
similar to that which occurs in the Sun-stars, only in this case the disc is raised 
entirely from the floor of the tank, and the whole movement is performed without any 
aid from the ambulacral feet; the latter, however, are kept in active motion during 
the whole of the righting movement. Sometimes only one arm, instead of two, is 
used as the fulcrum over which the disc and the other arms are thrown. In all cases 
