abound in his writings always remind me of certain cloud-forms which may sometimes be seen in the western sky, especially over horizons of the sea. They are often most glorious and imposing. Great lines of towers and of far-reaching battlements give the Impression at moments of mountainous solidity and strength. But as we gaze upon them with wonder, and as we fix upon them a closely attentive eye, the edges are seen to be as unsteady as at first they appeared to be enduring. If we attempt to draw them we find that they melt into each other, and that not a single outline is steady for a second. In a few minutes whole masses which had filled the eye with their majesty, and with impressions as of the everlasting hills, dissolve themselves into vapor and melt away.
Such are the cloud-castles which mount upon the intellectual horizon as we scan it in the representations of the mechanical philosophy. Nothing can be more fallacious than the habit of building up definitions out of words so vague and abstract that they may signify any one of a dozen different things, and the whole plausibility of which consists in the ambiguity of their meanings. It is a habit too which finds exercise in the alternate amusement of wiping out of words which have a definite and familiar sense, everything that constitutes their force and power. Let us take, for example, the word "function." There is no word, perhaps, applicable to our intellectual apprehensions of the organic world, which is more full of meaning, or of meaning which satisfies more thoroughly the many faculties concerned in the vision and description of its facts. The very idea of an organ is that of an apparatus for the doing of some definite work, which is its function. For the very reason of this richness and fullness of meaning, in this word conjoined with great precision, it is unfitted for use in the vapory cloud-castles of definition which are the boasted fortresses of ideas purely physical. And yet function is a word which it is most difficult to dispense with. The only alternative is to reduce it to some definition which wipes out all its special signification. Accordingly, Mr. Herbert Spencer has defined function as a word equivalent to the phrase "transformations of motion"[1]—a phrase perfectly vague, abstract, and equally applicable to function or to the destruction of it, to the processes of death or the processes of life, to the phenomena of heat, of light, or of electricity, and completely denuded of all the special meanings which respond to our perception of a whole class of special facts.
Of course the attempt breaks down completely to describe the facts of nature in words too vague for the purpose, or in words rendered sterile by artificial eliminations. It is not Dar-
- ↑ "Principles of Biology," vol. i, p. 4.
