40 J. S. HALDANE. learning he showed that his actions were really purposive ; and therefore it is a warrantable conclusion that his ap- parently mechanical obedience is at bottom a purposive obedience. Now if, in the case of the ganglion-cells con- cerned in reflex action, we find similar evidence of their behaviour being originally purposive, we shall similarly be warranted in concluding that reflex action is at bottom purposive and not mechanical. It will be sufficient to compare in respect of their reflex actions a mammal with one of the lower vertebrates, such as a frog ; as it may safely be inferred from analogy that reflex action in a frog corresponds roughly with an early stage in the development of reflex action in a mammal. If the spinal cord of a frog be severed from the brain, and a stimulating substance be applied to the skin of one of the hind limbs, the leg is moved in such a way as to make it evident that an attempt is being made to get rid of the source of irritation. And if the leg be held, the other leg is made use of to remove the irritating substance. The animal is also capable of performing many other actions which appear to be to a certain extent really purposive, and scarcely capable of being explained by the hypothesis of a prearranged mechanism in the cord. In the case, now, of a mammal whose cord has been similarly severed from the brain, what is observed is different. Complicated and highly co-ordinated reflex actions are, it is true, performed. But there is not the same power of adapting these movement to varied circumstances. In other words, the behaviour of the animal corresponds much more closely to what one would expect if the centres in the cord were only so many mechanisms, each previously arranged so as to act in a certain definite way. Now if it were true that reflex action, looked on as a purely mechanical phenomenon, is the type of the behaviour of the nervous system generally, one would expect to find this type manifested, not less, but more clearly in the case of the lower animals. But the opposite of this is found to be the case. Therefore the mechanical interpretation given to the facts of reflex action can be only superficially correct. The ganglion-cells of the cord can no more be regarded as mechanisms than can the ganglion-cells of the brain. As regards the doctrine often taught, that reflex action is the fundamental type to which the behaviour of the nervous system generally may be referred, physiologists would seem to have been led astray by an erroneous a priori conception. What is true of the ganglion-cells of the cord is true of the cells in parts of the body other than the nervous system.
