But if any one desire a fuller and more exhaustive treatment than even Maury gives of the hallucinations hypnagogiques, he had best read M. Baillarger's article "De Iinfluence de l'etat intermediate a la veille et au sommeil sur la production et la marche des hallucinations" in the Annates medico-psychologiques du systeme nerveux. Maury on this point says: "On voit, on songe (just before going to sleep) des personnes et des objets auxquels on n'avait pas pense depuis des annees" (when he says "on voit," he means, I suppose in the "mind's eye," for he specially says "devant mes paupieres fermees). And he goes on: "Ces apparitions sont la preuve que l'association plus ou moins logique des idees ne saurait expliquer tous les reves; car c'est le phenomene de l'hallucination hypnagogiques qui s'y poursuit." "La preuve," proof, is always a mighty big word. It means a deal. And with the highest respect to M. Maury it certainly seems too big a word for the kind of evidence he offers us here, just as, also, "car" seems to be a word of far more strict logical sequence than we can find justified by the force of the argument.

It does not in the least follow, because we cannot trace the association of ideas right on from their source to this or that phenomenon, or link in the chain, that the phenomenon was of necessity suggested by something outside that association; indeed the ordinary mind is rather surprised than otherwise when, suddenly asking itself the question, as often happens: "What in the world made me think of such or such a fact or person?" it succeeds in answering the question, in tracing the train of thought to its source. And if there be this difficulty in tracing the course of waking thought, how much more likely that there would be difficulty about tracing the much looser, irresponsible and uncontrolled suggestions of ideas during sleep, when the attention was scarcely alive at all, for the time being, to take notice of it.

The operation of the attention has so large an effect in determining sleep or wakefulness that no other mental faculty has so much. A very dull sermon or book that lets the attention flag is an excellent soporific. If we are wakeful, with the attention concentrated, despite ourselves, on some worry, the best means of dispelling the wakefulness is to read a book that shall distract the attention to another topic. The same function is intended to be performed by the sheep -personally I always find it more amusing to turn these hallucinations hypnagogiques into pigs - going one by one through a gate. The pigs rush and squeal more amusingly than the sheep, and, I think, divert the attention better. Concentration of mind, even the listening for an expected sound, is fatal to sleep. Absence of concentration is its condition. It is not a synonym of sleep, for after the attention falls into abeyance the semi-comatose condition of the senses has to supervene before one falls asleep. A very curious fact is related to me out of the experience of one of the many people who have been good enough to help me with instances about dreaming. It is to the effect that when the writer falls lightly and for a short time asleep, he always finds that his dreams are of events that have happened lately.

When on the contrary he is awakened from a sound sleep, he always finds his dream occupied with events that happened long ago, almost as if the thought in the sounder sleep went back to some of the deeper layers of reminiscence, to speak in this metaphorical way about it. It is a case that is the more worth noting because (were it found to be confirmed by others) it seems as if it might throw some new light on the many problems that still beset the whole inquiry. It does, however, require a good deal of corroboration, to make it of any value for the drawing of inferences; but if it should be found to be a typical condition in dreams, it must seem as if it had some connection with the tendency so often observed in the very old, whose intellectual powers are failing, to remember events that happened long ago far more clearly than the event of yesterday.

It is worth remark, too, and is not altogether alien from our subject, that the perception of time which is so altogether wanting in dreams depends almost entirely on the attention. The attention is in abeyance in dreams; and so, too, the perception of time. The typical case and historical instance is that of Lord Holland who fell asleep while a book was being read to him, awoke to hear the finish of the sentence that he had heard begun, and in the momentary interval dreamed a dream extending over a considerable time. The loss of perception of time is not of course restricted to a dream state, although it is in that condition that it is most constantly and completely exhibited. In minutes, or even hours, of waking reverie, not to speak of the trance and the like abnormal states, we lose all idea of time; and confess as much, more or less consciously, in common talk, when the clock strikes or the dinner bell sounds, to arouse us, and we ejaculate: "By love, I had no idea it was so late"!

It is more than probable that the whole conception of time, depending on the exercise of the attention, which is an exercise of the reasoning power, scarcely exists among the lowest nations of mankind, and is developed progressively with the development of the reasoning faculty. Evidently a complex state of society, requiring a nice fitting of the various duties of the day, fixed meal times and the rest, demands a far closer attention to the occupation of time than is required in the pastoral, nomadic or the hunter state.

An American, who had been much among the Indian tribes, described to me their indifference to time as something almost "divine," to use his own word. A band of the Pawnees he said, would sometimes encamp for a while in the neighbourhood of his home-stead on their way to a certain forest where they found suitable poles for their wigwams. And sometimes they would go on the next day, or sometimes they would stay for a month, or even three years (that was the longest period that he named to me) before moving on to complete the business on which they had set out Nor did they seem to be actuated by any motive for staying or for moving. No plan involving any concern with time seemed to have entered into their heads, unless it were to move on when their present wigwam poles should actually fall to pieces. My informant added that though he and all the neighbour-hood were on excellent terms with the Indians, they were greatly relieved when they did at length decide on a move, because they were, one and all, expert and unscrupulous horse-thieves. The whole story illustrates an indifference to time that may indeed be described as "divine," or equally well as "dreamlike ".

In a work, to dignify by such a name a book so entirely "popular" as this, it is not perhaps worth while to go deeper into the physiological side of sleep and dreaming. Without some little sketchy attempt of the kind it did not seem very easy to bring the reader up (or down) to that equal state of ignorance with the writer which is necessary if they are to understand each other in the chapters that follow. If any are brave enough to go further into the subject in the writers company, the latter believes he can promise them less dry and dismal entertainment in the pages to come.