medium-stock, and thinking how I might improve some item. This being the dramatically probable human way, I think differently of the whole type, taken collectively, from the way in which I may think of the single instance. I find myself believing that there is "something in" these never ending reports of physical phenomena, although I have n't yet the least positive notion of the something. It becomes to my mind simply a very worthy problem for investigation. Either I or the scientist is of course a fool, with our opposite views of probability here; and I only wish he might feel the liability, as cordially as I do, to pertain to both of us. I fear I look on Nature generally with more charitable eyes than his, though perhaps he would pause if he realized as I do, how vast the fraudulency is which inconsistency he must attribute to her. Nature is brutal enough, Heaven knows; but no one yet has held her non-human side to be _dishonest_, and even in the human sphere deliberate deceit is far rarer than the "classic" intellect, with its few and rigid categories, was ready to acknowledge. There is a hazy penumbra in us all where lying and delusion meet, where passion rules beliefs as well as conduct, and where the term "scoundrel" does not clear up everything to the depths as it did for our forefathers. The first automatic writing I ever saw was forty years ago. I unhesitatingly thought of it as deceit, although it contained vague elements of supernormal knowledge. Since then I have come to see in automatic writing one example of a department of human activity as vast as it is enigmatic. Every sort of person is liable to it, or to something equivalent to it; and whoever encourages it in himself finds himself personating someone else, either signing what he writes by fictitious name, or, spelling out, by ouija-board or table-tips, messages from the departed. Our subconscious region seems, as a rule, to be dominated either by a crazy "will to make-believe," or by some curious external force impelling us to personation. The first difference between the psychical researcher and the inexpert person is that the former realizes the commonness and typicality of the phenomenon here, while the latter, less informed, thinks it so rare as to be unworthy of attention. _I wish to go on record for the commonness_. The next thing I wish to go on record for is _the presence_, in the midst of all the humbug, _of really supernormal knowledge_. By this I mean knowledge that cannot be traced to the ordinary sources of information--the senses namely, of the automatist. In really strong mediums this knowledge seems to be abundant, though it is usually spotty, capricious and unconnected. Really strong mediums are rarities; but when one starts with them and works downwards into less brilliant regions of the automatic life, one tends to interpret many slight but odd coincidences with truth as possibly rudimentary forms of this kind of knowledge.