remembers from day to day the smell of my clothes; he recognises old friends after long absence by their odour; he recollects and knows the distinctive perfumes of every bird or animal. Nay, more, it is probable that these smell-memories are consolidated into a regular succession in his mind, just as sight-memories are consolidated in ours. If you and I have once been to a place, we find our way back again by remembering the visible aspect of the road, the various streets and turnings, the trees and houses, the hills and valleys. But if Anacharsis has once been to a place and goes there again, you will see him taking notes as he runs along, not with his eyes, but with his nose. You will see him give a hearty whiff of recognition at every corner, or take a dubitative long breath at an uncertain cross-road. It has long been known that dogs conveyed by train to a strange place, or else carried in a covered basket, have often found their way home again at once and without difficulty. Now, Mr. A. R. Wallace suggests that they probably do so by observing and remembering the smells they have met with on the way; and Professor Robertson further points out that such memory is the less remarkable when we recollect that the sense of smell in dogs is most likely an unbroken whole. "The dog's world," he says, "may be, in the main, a world of sights and smells continuous in space." In other words, while you and I think of a given field as a mass of visible objects, Anacharsis very probably thinks of it as a mass of smells. Most likely it seems no more remarkable to a dog that he can remember a whole string of odours in their regular order than it seems remarkable to us that we can remember our way from Hyde Park Corner to Oxford Circus by means of a whole string of visible objects, observed and recollected as signs of the road.

Again, when the dog thinks of anything, its smell must be a main part of his thinking about it. He must remember a man always to a great extent as a smellable thing. Indeed, the dog even dreams about smells, as we may see by his sniffing and growling in his sleep. If you watch him narrowly, you can notice that at one time he seems to dream of hunting, puts his nose down against the hearth-rug, and draws in his breath with a kind of quiet satisfaction, as if engaged in silently tracking down his game; while at other times, he appears to dream about an enemy, when he may be observed to take sharp snorts of a convulsive kind, and to yelp angrily as he raises his head a little from the ground, in the half-assumed attitude of battle.

These examples lead us on to the fact that smells must also be largely connected in the canine mind with all kinds of appropriate emotions. Some of them must rouse associated feelings of devotion

to a master, of affection, of anger, of dislike, of excitement, or of fear. The least odour of rat or rabbit will set a terrier frantic with the hunting fever: the spoor of a negro will drive the bloodhound wild with the instinct of tracking down the fugitive. I have known many Cuban bloodhounds in Jamaica which always fawned upon a white man, friend or stranger, but could not be trusted for a moment by any black man, including even the servants who ordinarily fed them. That scent, not colour, formed the means of discrimination is certain, for they attacked negroes at night even more than by day. Everybody must have noticed thousands of similar instances, where particular emotions were obviously associated in the minds of dogs with particular odours.

Even in our human brains, with their very shrivelled olfactory lobes, such emotional and intellectual associations with perfumes occasionally occur. We have all observed that now and then an odour recalls some half-forgotten scene or some faint wave of feeling, such as tenderness or vague melancholy. It is even usual to speak of smell as being a sense exceptionally apt so to recall ideas or emotions. But the exact contrary is really the truth. We notice these cases just because of their extreme rarity. Nobody would think of remarking it as a curiosity that a certain visible or audible object recalled another; nobody would dream of saying anything so obvious as that the sight of their mother's face or the sound of their sister's voice vividly aroused pleasant memories and associations. But on the rare occasions when a smell faintly calls back an idea or a feeling, we are struck with the unusualness of the effect, and so make a mental note of it. Thus, the mere oddity of the experience stamps it on the mind, and induces people who are unaccustomed to psychological analysis to jump at the conclusion that smell is peculiarly powerful in recalling associated notions; whereas the exact opposite is really the truth, at least as regards the human race. Sight, touch, and hearing are with us the leading intellectual senses; the senses, that is to say, which have the most numerous and most definite connections between themselves, as well as with the other senses, and which, therefore, most vividly call up associated ideas. But these rare smell-currents, these trains of thought initiated by an odour, are nevertheless extremely interesting, because they enable us dimly to realise how the sense of smell acts in the lower animals. They may be regarded as survivals of the old nervous connections, now almost obliterated in our brains. In the same way we know that many idiots-human beings who have hardly developed beyond the brutal stage-are in the habit of smelling at food and other objects given to

them; and this would seem to be a similar survival from an earlier state. Smell is also said to be a much more important endowment amongst some savages than in civilised races. Unfortunately, I do not know whether in the brains of such idiots or savages any special note has ever been taken of the relative development of the olfactory lobes.

I hope, however, that it is now clear why, on the one hand, the central organs of smell are so large in the dog; and why, on the other hand, he has been enabled to develop so high a degree of sagacity in spite of his total lack of delicate tactual or grasping organs. Smell, as we have seen, not only supplements sight and supersedes touch with him, but also forms endless lateral connections in every direction, so as to modify his whole conception of the universe. And since he does not manipulate things for purposes of inanufacture, as we do, but merely eats, tears, or hunts them, smell really proves just as useful to him as touch does to us. Being itself, as Mr. Herbert Spencer says, an "anticipatory taste," it is well fitted for the final court of appeal in cognising external objects in the case of a carnivorous animal, which uses its mouth, jaws, and teeth as its only substitute for human implements. So that the dog's intellect and the dog's senses are on the whole admirably adapted to just the sort of life which the dog must necessarily lead.

Indeed, one may

Of course many animals besides dogs have a very developed sense of smell. Dr. Bastian notes, amongst others, the American bison, in whom it is so keen that neither men nor dogs can approach him except from the leeward side; and the camel, which is said to discover water in the desert at a distance of a mile by means of sniffing. He also notices the well-known case of the deer, whose delicacy of scent is familiar to all Highland stalkers. say roughly that an acute and discriminative sense of smell is indispensable to all the carnivores for tracking their prey, and to all the ruminants for escaping their enemies. Horses, likewise, display the same high powers of scent in a remarkable degree, and with them the nose, doubtless, largely supplements their tactile and mobile upper lip. Mr. Darwin mentions the case of a blind mare in a stagecoach who regularly pulled up at certain points of the journey, opposite public-houses and other recognised stopping-places, which she seemed to distinguish by her nose quite as well as other horses did with their eyes. A frightened horse may often be reassured by making him smell the object at which he shied: he then learns what sort of thing it really was. But amongst still lower creatures it is probable that smell plays even a larger relative part than in the

mammalian races. With fishes it apparently forms the most im portant sense of all, guiding them to their prey from immense distances. Anglers know that trout will often refuse artificial flies if quite scentless, but will eagerly dart at them when they have been gently smeared with a piece of worm or a bit of the real insect whose form and colour they imitate. And in insects generally, smell seems in no way less valuable than sight as a guiding and directing agency.

Ants, however, present us with the most curious and perfect example of all; and though their intelligence may seem at first sight to have little relation with the universe of dogs, I think we shall see that they do really cast a great deal of indirect light upon the canine mind. There are a few insects which possess in their heads a mass of nervous matter that may be fairly considered as analogous to the brain of vertebrate animals. These insects are the bees, flies, and ants. As a rule, the nervous system of articulate animals is very scattered, consisting of several disjointed ganglia distributed pretty equally amongst the various segments of the body. But in these higher races the head contains a small mass of higher co-ordinating centres, superimposed upon the ganglia in direct connection with the sense-organs; and this mass has functions apparently similar to those of our own brains. Now, in the bee, the tiny brain in question must obviously be engaged in correlating and co-ordinating sights and smells with motions. The bee has a developed eye, with which it perceives the forms and colours of flowers; and it also has a developed organ of scent, with which it perceives the perfumes of thyme or marjoram; and it governs the movements of its wings, legs, and mouth in accordance with the information thus given it. But the ant, which is a near relative of the bee, has lost its wings (at least, in the case of the neuters), and has taken to a life of running about on its six legs instead of flying; a change which is correlated with its carnivorous habits, just as the structure of the bee is wholly dependent upon its honey-sucking propensities. Under these circumstances the ant has almost lost its eyes, which now survive only in the winged males and females, while the workers are almost, or in some species entirely, blind. To slow and wingless carnivorous creatures scent seems to prove more useful than sight. At any rate, while the ants have quite got rid of their eyes, for all practical purposes, they have developed their sense of smell to such an extent that it serves as their one and only intellectual monitor. Since ants are wholly devoid of hearing, it appears that the whole raw material of their intelligence, the single set of sensations upon which their little brains can work, is given them by odours. What touch is to the

blind man, that is scent to the almost blind ant. They smell their way from place to place; they recollect the road to their nest by smell; they recognise friends and enemies by means of scent; they track their path through life by olfactory sensations alone. Their example shows us how high an intelligence may be evolved from the constant use of this one sense in isolation.

Now, we may fairly say that in this particular the dog stands, as it were, half-way between ourselves and the ant, with one point of sensuous superiority to each of us. In man the sense of smell has become a mere relic, of no practical or intellectual importance. We may very occasionally sniff at a bottle to discover what are its contents; but as a rule our whole conduct in life is guided by sights and sounds alone. With the ant, on the other hand, the sense of sight has become a mere relic, as unimportant to his life at large as smell is to our own. But with the bee and the dog both sight and smell are intellectual senses of the first order, guiding and directing their motions every moment of their lives. While man's world is mainly a world of sights and touches, and while the ant's world is mainly a world of smells, the dog's world is mainly a world of sights and smells combined, with an occasional interruption of sounds, touches, tastes, and internal feelings.

Another insect analogy may further help us to the comprehension of yet a more difficult problem in dog psychology. If I take an example from Dr. Bastian, I shall make the nature of the problem clearer to my readers. A hound was sent, he says, from a place in County Dublin to another in County Meath, and thence, long afterwards, conveyed to Dublin town. There he broke loose and made his way back at once to the kennel in his first home, thus completing the third side of a triangle by a way which he had never travelled in his life. From this and many similar circumstances, Dr. Bastian concludes that the lower animals may, in some cases, possess what he calls a "sense of direction." Now, I am myself averse to such somewhat mystical explanations of half unknown and half uncertain facts as that involved in the hypothesis of a seventh sense. It savours a little too much of the method by which we have been deluged with spiritualism, animal magnetism, psychic force, and a vast number of like unprovable entities. I prefer to look for an explanation of the facts, if facts they really are, among better known and undoubted realities. It so happens that we have analogies at hand which amply suffice to cover the cases in point. We have seen already that both the deer and the bison are extremely sensitive to distant smells wafted by the wind, and that it is impossible to