It is fairly easy to test the sense of smell in these creatures, the only necessary precaution being that no acid or pungent substances capable of having an irritating effect on the skin, such as vinegar or ammonia, must be employed. Such perfumes as lavender or heliotrope are free from this defect. Take a clean glass rod and present it to the spider as before, and no notice is taken. Now dip it in oil of lavender, allow it to dry, and present it again. Most spiders respond to such a test, Epeirids generally raising the abdomen, and rubbing one or other of the legs against the jaws, while jumping spiders generally raise the head and back away from the rod. Different essences produce different effects, but there is seldom any doubt that the creature is aware of their presence; it is not deficient in the sense of smell, but its localisation has hitherto baffled research. The sense of taste does not seem to have been made the subject of any definite experiments among spiders, though such experiments might well lead to interesting conclusions, and the reader might do worse than undertake some on his own account. It would be easy, for instance, to supply a garden-spider with various insects which are generally rejected by other insectivorous animals, and to note its behaviour. It might refuse to have anything to do with them, or it might sample them and turn away in disgust. In the first case the explanation might be that it was warned of their probably evil taste by their coloration or smell, but in any case here is an interesting little field for research. It is the general belief among arachnologists that the sense of taste is well developed among spiders, and it is highly improbable that a sense so necessary for the discrimination of suitable food should be lacking in animals with so respectable a sensory equipment. There is no doubt at all that the sense of touch is extremely well developed in spiders, especially perhaps, in the sedentary groups, and it is probable that, under ordinary circumstances, the garden-spider works almost entirely by its guidance. Whether in the centre of the web or in its retreat under a neighbouring leaf it is in direct communication with every part of its snare by silken lines, and the least disturbance usually suffices to bring it to the spot; and then, as we have said, it will generally touch the disturbing object, however unpromising in appearance, before deciding on its line of action. There is little doubt that many of the numerous hairs and bristles with which its limbs are furnished are distinctly sensory in function.