At all events, all reserves made for fraud, which I recognise possible though improbable, chanting or singing in common has appeared to me to have a favourable influence on the phenomena. I have had occasion of verifying this effect of intoned words; I am unable to give its explanation, although we may suspect what it is likely to be. I will simply recall to mind the rôle which intoning or singing plays in religious ceremonies and in magical operations: the words ‘incantations,’ ‘enchantments,’ are very significative, from that point of view. The erudite will remember the magic songs of the 11th eclogue of Theocritus, and of the 8th of Virgil. The Hindoo magicians intone their mentrams. Nothing is more widespread than this belief in the supernatural virtue of singing, of the cadenced and modulated word. As the supernormal facts which I relate appear to me to have been known from the earliest times—however ill-interpreted they may have been—I am inclined to believe, that the superstitions relative to the magical power of song are not without a foundation of truth. This appears most improbable, and no one is more astonished than myself, to find myself admitting this possibility. I admit it nevertheless. I am inclined to think, that the greater part of popular beliefs have some foundation; the particle of truth which they contain is often very feeble, because ignorance, fear, imagination mask it under accessory and unreasonable beliefs, which smother it. There would be many interesting analogies to point out on this subject, if I had not systematically forbidden myself all manner of theoretical commentary. All the same, I will remark that the most worthy spiritists recommend singing or music during seances. I will cease, for I can only repeat here the considerations which I have already presented concerning the relation between the nervous energy, whatever it may be, and luminous phenomena; the connection appears to be very close indeed. The physiognomical aspect of these phenomena is similar to that of sonorous and motor phenomena: It tends to personification, and it is probable, that imperfect luminous forms are but rude outlines of a real form. That form is not always human, although it appears to be so as a rule. I have given examples, where the appearance was that of an animal or of an object. I have never been able to converse with the form itself, when it was human; but I have experimented with mediums who thought they conversed with the forms. These all claim to be the spirits of deceased persons. What renders this unanimity particularly interesting is that one of the mediums, with whom I have observed the finest phenomena of human appearances, is by no means a spiritist.