Now, this theory of conscious automatism is not merely a legitimate outcome of the theory that nervous changes are the causes of mental changes, but it is logically the only possible outcome. Nor do I see any way in which this theory can be fought on grounds of physiology. If we persist in regarding the association between brain and thought exclusively from a physiological point of view, we must of necessity be materialists. Further, so far as we are physiologists our materialism can do us no harm. On the contrary, it is to us of the utmost service, as at once the simplest physiological explanation of facts already known, and the best working hypothesis to guide us in our further researches. But it does not follow from this that the theory of materialism is true. The bells of St. Mary's over the way always ring for a quarter of an hour before the University sermon; yet the ringing of the bells is not the cause of the sermon, although, as long as the association remains constant, there would be no harm in assuming, for any practical purposes, that it is so. But just as we should be wrong in concluding, if we did not happen to know so much about the matter as we do, that the University sermon is produced by the vibration of bells in the tower of St. Mary's Church, so we may be similarly wrong if we were definitely to conclude that the sermon is produced by the vibration of a number of little nerve-cells in the brain of the preacher. Now, if time permitted, and if I supposed that you would all care to go with me into matters of some abstruseness, I could certainly prove that whatever the connexion between body and mind may be, we have the best possible reasons for concluding that it is not a causal connexion. These reasons are, of course, extra-physiological; but they are not on this account less conclusive. Within the limits of a lecture, however, I can only undertake to give an outline sketch of what I take to be the overwhelming argument against materialism.