Another experiment of the same kind showed like the others that it was the sound of the word and not the idea it conveyed which was perceived by the brain. Then the words _chandelle_, _haridelle_, were pronounced many times in rapid succession in his ear. He awoke suddenly, saying to himself, _c’est elle_. It was impossible for him to recall what idea he had attached to this dream. 8th Experiment. A drop of water was allowed to fall on his forehead. He dreamed that he was in Italy, that he was very warm, and that he was drinking the wine of Orvieto. 9th Experiment. A light, surrounded with a piece of red paper, was repeatedly placed before his eyes. He dreamed of a tempest and lightning, which suggested the remembrance of a storm he had encountered in the English Channel in going from Merlaix to Havre. These observations are very instructive. They show conclusively that one very important class of our dreams is due to our bodily sensations. I have frequently performed analogous experiments on others, and had them practiced on myself, and have rarely failed in obtaining decided results. They strongly inculcate the truth of the conclusions arrived at in the foregoing chapter, and they serve as important data in enabling us to understand the division of the subject next to be considered. In regard to the immediate cause of dreams the opinions of authors are very diverse. The older writers ascribe them to the rise of vapors from the stomach, to the visitation of demons, and other fanciful causes. Bishop Bull[84] declares that he knows from his own experience that dreams are to be ascribed “to the ministry of those invisible instruments of God’s providence that guide and govern our affairs and concerns, viz., the angels of God;” and Bishop Ken held a similar view.