To sum up, then, and to conclude. I would ask you to see the connection between words and ideas as in the first instance arbitrary. No doubt in some cases an imitation of the cry of some bird or wild beast would suggest the name that should be attached to it; occasionally the sound of an operation such as grinding may have influenced the choice of the letters g, r, as the root of many words that denote a grinding, grating, grasping, crushing action; but I understand that the number of words due to direct imitation is comparatively few in number, and that they have been mainly coined as the result of connections so far-fetched and fanciful as to amount practically to no connection at all. Once chosen, however, they were adhered to for a considerable time among the dwellers in any given place, so as to become acknowledged as the vulgar tongue, and raise readily in the mind of the inhabitants of that place the ideas with which they had been artificially associated. As regards our being able to think and reason without words, the Duke of Argyll has put the matter as soundly as I have yet seen it stated. "It seems to me," he wrote, "quite certain that we can and do constantly think of things without thinking of any sound or word as designating them. Language seems to me to be necessary for the progress of thought, but not at all for the mere act of thinking. It is a product of thought, an expression of it, a vehicle for the communication of it, and an embodiment which is essential to its growth and continuity; but it seems to me altogether erroneous to regard it as an inseparable part of cogitation." The following passages, again, are quoted from Sir William Hamilton in Professor Max Muller's own book, with so much approval as to lead one to suppose that the differences between himself and his opponents are in reality less than he believes them to be. "Language," says Sir W. Hamilton, "is the attribution of signs to our cognitions of things. But as a cognition must have already been there before it could receive a sign, consequently that knowledge which is denoted by the formation and application of a word must have preceded the symbol that denotes it. A sign, however, is necessary to give stability to our intellectual progress--to establish each step in our advance as a new starting-point for our advance to another beyond. A country may be overrun by an armed host, but it is only conquered by the establishment of fortresses. Words are the fortresses of thought. They enable us to realize our dominion over what we have already overrun in thought; to make every intellectual conquest the base of operations for others still beyond." "This," says Professor Max Muller, "is a most happy illustration," and he proceeds to quote the following, also from Sir William Hamilton, which he declares to be even happier still.