Not even a horny layer of dead cells is, however, always sufficient protection, and the growing layer has sometimes to supplement it by hair or feathers. How hair is developed is shown in the accompanying diagram (50). The growing layer sends a strand straight downwards into the connective tissue, which forms the basement of the skin. The cells in the middle of this strand, which behaves like ordinary skin, are the least well nourished, and accordingly die and leave a tube. This tube, if no further development took place, might become a sweat gland; but if it is to give rise to a hair it becomes cup-shaped at the base, enclosing a small loop of bloodvessel. The cells just above the capillary, being better nourished than the rest, grow more rapidly than their neighbours, and the result is that a column of cells which we know as a hair pushes its way up through the tube. (See Diagram 50.) This outer layer comes everywhere between the main bulk of the body and the outer world. Hair and sweat glands do not by any means represent its only modifications. Teeth are formed from it in somewhat the same manner as hair, while we have already seen that it gives rise to the whole nervous system. The next thing which we have to consider is how knowledge of the external world reaches the central nervous system. Sensations of touch, temperature, and pain are fairly easy to understand, since the nerves which convey such impressions have numerous endings in the skin. End organs of nerves in the joints and muscles doubtless enable the animal to perceive and estimate strain and resistance in moving or lifting things. But the power of perceiving the chemical peculiarities of things; light, involving the formation of visual images, which we call seeing; sound; and position and equilibrium, it is not possible for the whole surface of the body to possess. The principle of division of labour is extended to the task of perception as well as to that of motion; and cells, with their property of responding to light, vibration, chemical stimulation, etc., are grouped together to form special organs, connected with the central nervous system by special nerves. Perhaps the most important factor which can influence protoplasm is the chemical nature of its surroundings; and in the first essay, on the general nature of protoplasm, we touched upon the way in which it is drawn towards some substances, and repelled by others.