44 J. S. HALDANE. cells to which they lead adapt themselves to the new desti- nation of their fibrils. In any case, the restoration of the function of the nerve is a very significant process ; and it would be a manifest abstraction from the actual facts to regard this process merely as the joint effect of keeping the cut ends in apposition and other local causes. Sewing together the cut ends has a therapeutic function, but cannot be regarded merely as a cause acting from without on the organism. Such considerations as these justify the conclusion that the scepticism above mentioned with regard to thera- peutics is based on an inadequate conception of the real nature of this art ; and that advances in the treatment of disease are likely to be as great as, if not greater than, advances in the ' diagnosis ' and ' prognosis '. Perception, in so far as it is treated of by physiology, is commonly assumed to be essentially nothing more than a mechanical process, in which physical impulses are trans- mitted from surrounding objects to the nerve-cells of the brain. The impressions thus produced may set a-going nervous processes of various kinids ; and these may end in muscular motion, or in other ways. It is unnecessary to consider here the doctrine that changes in the nervous system give rise to, or are accompanied by, consciousness. That doctrine in its various forms has been exhaustively and finally dealt with by Green. I shall only consider the physical, or what is often called the ' objective,' side of per- ception that is to say, perception in so far as it is a process which may be investigated by physiological methods. The use of the word ' perception ' in this sense is no doubt objection- able ; but it is hard to find a more suitable expression. If it can be shown that in perception the surroundings affect the organism in a purposive manner, it will be possible to draw from this fact the same inferences as were previously drawn from the fact that the organism acts purposively on the surroundings. A condition which is more or less essential to perception is attention on the part of the perceiv- ing organism. Now this of course might be taken to mean that certain causes acting from within the bodily mechanism must co-operate with those acting from without in order that there may be perception. But to say this is only to put aside, and not to solve, what is a real difficulty in the way of a mechanical interpretation. For it may be asked what it is that causes attention to be directed to one object rather than another. The ultimate answer to this question is that those objects are attended to, the perception of which is of advantage to the perceiving organism, or conduces in some way to its life. And as attention is directed in this manner,
