Similar reflexes are found in the other groups of echinoderms, particularly among the starfishes and brittle-stars. These are able, by reflexes, to free themselves from all sorts of difficult situations. Preyer fastened Asterias by sticking five nails with big heads into a board close to the angles between the five arms of the starfish, but without injuring the latter. The confined animals, which could not escape upwards owing to the large heads of the nails, freed themselves in various manners by squeezing, bending and turning the body now in this, now in that way (fig. 2). This they did, not according to any definite plan, nor from any insight into their position or of the necessary steps to be taken to attain freedom, but solely as a result of their physiological consti tution and of the various possibilities of movement at their dis position. Brittle stars, when an india-rubber tube has been passed over one of the five arms, escape very quickly by propping the neighbouring arms against the rubber and so pushing it off. A starfish turned on its back first extends all its tube-feet which then feel around as if seeking for a solid object on which to attach themselves. Then several of the tips of the arms turn the right way up. So far all this takes place in no definite order. Apparently each part works independently of the others. Only after a period of indecision, a co-ordination of the movements is brought about through the radial nerves. The tube-feet of one or two of the arms which have turned over (the directive arms) attach themselves to the sub-stratum, whereupon the body with the re maining arms turns over. This directive influence assumed by one or two of the arms consists in the inhibition by impulses from the directives of the independence of the remaining arms. Something similar occurs when the starfish frees itself from the pentagon of nails. Now, it is both interesting and important that an amputated arm of a starfish, or indeed any isolated part of an echinoderm, behaves to stimuli as if it were still connected to the whole animal. It does this so long as its nervous reflex paths are un injured. This is the real proof of the purely reflex character of the different reactions observed in the normal animal.
Hermit-crabs hide their unprotected naked abdomen in the shells of sea-snails which, as occasion offers, they exchange for more comfortable shells. Before slipping into such a new shell, they first put their claws inside, as though they wished to begin by investigating the new abode. Conical snail-shells are held in such a manner that the crab's telson to a certain extent holds the tip of the house and the crab, standing almost on its head, turns the shell round. This appears eminently purposeful, for by this means small stones and grains of sand which may happen to be inside the shell fall out. If these were not removed they might injure the soft hind-parts of the crab. If a number of empty shells of sea-snails are placed in an aquarium, containing hermit crabs, the animals change their habitations continuously. As soon as a crab happens upon an empty shell, it feels it over and then steps in. The fact that frequently in this pursuit a shell which suits the crab admirably is exchanged for another obviously smaller and less suitable, shows plainly that a purely reflex action is involved. Bohn expresses himself thus : "The animal carries out a quite definite series of movements, from the moment when it touches the shell until the instant when it hides its abdomen in it. Each movement causes the next one to follow on. Hence it is possible, to a certain extent, to cheat the crabs. If one places them upon a round stone or wooden ball, then they turn this round in all directions, press their bodies against it and seek for the mouth, which of course does not exist. They look for
the opening likewise in an unusually convex oyster shell." Thus the hermit-crabs, too, behave in this respect like automata. The stimulus supplied by the snail-shell, acting through the appro priate reflex paths, mechanically causes the movements just described. And this seems just the same when the movements are unnecessary or even quite meaningless.
A peculiar type of reflex is found in many animals, which under certain circumstances serves as a defence when the animals are seized by an enemy. This is the autotomy or self-mutilation reflex. By it animals are able to throw off the parts of the body or limbs which have been caught, so that with this loss they can still escape. In this way many annelid worms throw off certain parts of the body, starfishes and brittle-stars reject their arms. Crus tacea, especially shore-crabs and lobsters, break off legs and claws which are caught and held. Spiders and many insects behave in a like manner. Since the place at which the limbs break off is usually suited morphologically to this purpose, a definite adaptation is involved.
A good example of behaviour made up almost wholly of re flexes is afforded by the ant-lion, the larva of Myrmeleo formi carius. These animals construct funnel-shaped hollows in sandy soil and dig themselves in up to their heads at the bottom of the funnel. Then they lie in wait for passing ants. When an ant approaches the margin of the funnel, the ant-lion, in response to the stimulus supplied by the sand-grains rolling down the walls of the funnel, performs the jerk-reflex; i.e., it throws its head upwards and backwards with a mighty jerk. Consequently the surrounding fine sand is thrown up and causes the ant, already advancing with difficulty in the heavy soil, to slip completely down. The jerk-reflex may be repeated a couple of times if further sand particles continue to fall down. If the ant chances to touch the inner side of the widely-open mandibles of its enemy, the snap-reflex follows. This involves the closing of the sharp jaws to seize the prey. This snap-reflex can be induced artificially by the touch of a fine needle. The construction of the funnel, too, is said to be performed reflexly. Such reflexes are very wide spread among insects. The larvae of dragon-flies, for example, live in water as predaceous animals. They catch their prey with their so-called mask. This is the modified underlip, consisting of pincers which can be projected rapidly for a considerable distance in front of the head. Since in this case optical stimuli bring the mechanism of snapping into action, the movement to and fro of any small object about the size of the normal prey is sufficient to cause these larvae to snap at them. If one lightly strokes the wing of a fly, Eristalis tenax, once or twice with a thread, the insect raises the hind-leg of that side and wipes the wing-surface with it. The numerous cleaning movements of insects come about, too, in a purely reflex way. As in echinoderms, amputated por tions of the body behave just as when they still formed a part of the whole animal, provided that the necessary nervous connec tions are still present. The stinging apparatus of the honey-bee, which remains fixed in the human skin after stinging owing to its backturned hooks, and is thus torn out of the abdomen of the bee, continues to bore even deeper into the wound. And it is common knowledge that one must beware of the hinder portion of a freshly killed wasp, even when this part of the body is cut off, for it continues for quite a time to execute stinging movements. The trunk portions of ants, after the head and hind part of the body have been cut off, continue for a half hour or an hour to walk around haltingly with short steps and rise up again if they fall over.