explained, this seems to clear the communications. Any means that can be adopted to render clearer the mind of the communicator, on the one hand, or improve the condition of the nervous mechanism of the medium on the other, should therefore be of great utility and should at least be tried. This being so, I now come to the heart of the matter, and offer a suggestion which, if followed out, might improve the physical body of the medium, and hence render the conditions better from _this_ side--as the presentation of objects might be supposed to render the conditions better from the other side. "I have pointed out before that, in certain cases, when it is desirable to restore the consciousness and to render its renewal more certain and clear (after an accident, e.g., that has knocked a person senseless) a mixture of oxygen gas is sometimes administered to the patient in order to produce these results. This being so, I ask: why may it not be a good idea to administer a diluted mixture of this gas to the medium when she is in a trance state--and when a communicator is attempting to convey his thought to the sitter by means of automatic writing? Might not such an experiment be tried, since no _harm_ could come to the medium if the oxygen were diluted and only sufficiently strong to effect the desired results? And might not its administration tend to improve the tone of the nervous system _pro tem._, and render clearer the consciousness that is trying to use it and manifest through it--just as one's own consciousness might be rendered clearer by the same device? Of course such a process might have the effect (especially at first) of breaking the trance altogether, and of reviving the medium. But if the medium understood the experiment beforehand, and the process were also explained to the controls, it is reasonable to suppose that--after some trials at any rate--the trance would not be broken, and that better, clearer results would follow. At all events, when some of our physicians in America are experimenting upon the effects of various electrical rays upon mediums in a trance, might not this far simpler and better-understood method be tried with more or less impunity? I at least suggest that it be so tried."