expression, in comparison with ordinary language, as going on one's hands and knees is in comparison with walking, or as walking compared with going by train; but it is as great an abuse of words to limit the word "language" to mere words written or spoken, as it would be to limit the idea of a locomotive to a railway engine. This may indeed pass in ordinary conversation, where so much must be suppressed if talk is to be got through at all, but it is intolerable when we are inquiring about the relations between thought and words. To do so is to let words become as it were the masters of thought, on the ground that the fact of their being only its servants and appendages is so obvious that it is generally allowed to go without saying. If all that Professor Max Muller means to say is, that no animal but man commands an articulate language, with verbs and nouns, or is ever likely to command one (and I question whether in reality he means much more than this), no one will differ from him. No dog or elephant has one word for bread, another for meat, and another for water. Yet, when we watch a cat or dog dreaming, as they often evidently do, can we doubt that the dream is accompanied by a mental image of the thing that is dreamed of, much like what we experience in dreams ourselves, and much doubtless like the mental images which must have passed through the mind of my deaf and dumb waiter? If they have mental images in sleep, can we doubt that waking, also, they picture things before their mind's eyes, and see them much as we do--too vaguely indeed to admit of our thinking that we actually see the objects themselves, but definitely enough for us to be able to recognise the idea or object of which we are thinking, and to connect it with any other idea, object, or sign that we may think appropriate?