From the dangers of hypothesis let us now turn to the dangers of no hypothesis. Little recognized as is the fact, it is nevertheless true that we cannot make the commonest observation correctly without beforehand having some notion of what we are to observe. You are asked to listen to a faint sound, and you find that without a pre-conception of the _kind_ of sound you are to hear, you cannot hear it. Provided that it is not strong, an unusual flavour in your food may pass quite unperceived, unless some one draws attention to it, when you taste it distinctly. After knowing him for years, you shall suddenly discover that your friend’s nose is slightly awry, and wonder that you never remarked it before. Still more striking becomes this inability when the facts to be observed are complex. Of a hundred people who listen to the dying vibrations of a church bell, almost all fail to perceive the harmonics, and assert the sound to be simple. Scarcely any one who has not practised drawing, sees, when in the street, that all the horizontal lines in the walls, windows, shutters, roofs, seem to converge to one point in the distance: a fact which, after a few lessons in perspective, becomes visible enough. Perhaps I cannot more clearly illustrate this necessity for {165} hypothesis as a condition to accurate perception, than by narrating a portion of my own experience relative to the colours of shadows. Indian ink was the pigment which, during boyhood, I invariably used for shading. Ask any one who has received no culture in art, or who has given no thought to it, of what colour a shadow is, and the unhesitating reply will be—black. This is uniformly the creed of the uninitiated; and in this creed I undoubtingly remained till about eighteen. Happening, at that age, to come much in contact with an amateur artist, I was told, to my great surprise, that shadows are not black but of a neutral tint. This, to me, novel doctrine, I strenuously resisted. I have a pretty distinct recollection of denying it point blank, and quoting all my experience in support of the denial. I remember, too, that the controversy lasted over a considerable period; and that it was only after my friend had repeatedly drawn my attention to instances in Nature, that I finally gave in. Though I must previously have seen myriads of shadows, yet in consequence of the fact that very generally the tint approaches to black, I had been unable, in the absence of hypothesis, to perceive that in many cases it is distinctly not black.