IV.

Much stress is usually laid, in accounts of the Gwyllgi, on the terror with which it inspires domestic animals. This confidence in the ability of the brute creation to detect the presence of a spirit, is a common superstition everywhere. An American journal lately gave an account of an apparition seen in Indiana, whose ghostly character was considered by the witnesses to be proven by the terror of horses which saw it. They were drawing the carriage in which drove the persons to whom the ghost appeared, and they shied from the road at sight of it, becoming unmanageable. The spectre soon dissolved in thin air and vanished, when the horses instantly became tractable. In Wales it is thought that horses have peculiarly this 'gift' of seeing spectres. Carriage horses have been known to display every sign of the utmost terror, when the occupants of the carriage could see no cause for fright; and in such cases a funeral is expected to pass there before long, bearing to his grave some person not dead at the time of the horses' fright. These phenomena are certainly extremely interesting, and well calculated to 'bid us pause,' though not, perhaps, for the purpose of considering whether a horse's eye can receive an image which the human retina fails to accept. Much weight will not be given to the fright of the lower animals, I fear, by any thoughtful person who has witnessed the terror of a horse at sight of a flapping shirt on a clothes-line, or that hideous monster a railway engine. Andrew Jackson Davis has a theory that we all bear about us an atmosphere, pleasing or repulsive, which can be detected by horses, dogs, and spiritual 'mediums;' this _aura_, being spiritual, surrounds us without our will or wish, goes where we go, but does not die when we die, and is the means by which a bloodhound tracks a slave, or a fond dog finds its master. Without denying the possibility of this theory, I must record that in my observation a dog has been found to smell his master most successfully when that master was most in need of a bath and a change of linen. Also, that when the master leaves off his coat he clearly leaves--if a dog's conduct be evidence--a part of his _aura_ with it. More worthy of serious attention is August Comte's suggestion that dogs and some other animals are perhaps capable of forming fetichistic notions. That dogs accredit inanimate objects with volition, to a certain extent, I am quite convinced. The thing which constitutes knowledge, in dogs as in human beings--that is to say, thought, organised by experience--corrects this tendency in animals as they grow older, precisely as it corrects the false conclusions of children, though never to the same extent. That a dog can think, I suppose no well-informed person doubts in these days.