Observation of the determinants which favour lapses in reading gives rise to doubt, which I do not like to leave unmentioned, because I am of the opinion that it may become the starting-point of a fruitful investigation. It is a familiar fact that in reading aloud the attention of the reader often wanders from the text and is directed toward his own thoughts. The results of this deviation of attention are often such that when interrupted and questioned he cannot even state what he has read. In other words, he has read automatically, although the reading was nearly always correct. I do not think that such conditions favour any noticeable increase in the mistakes. We are accustomed to assume concerning a whole series of functions that they are most precisely performed when done automatically, with scarcely any conscious attention. This argues that the conditions governing attention in mistakes in speaking, writing, and reading must be differently determined than assumed by Wundt (cessation or diminution of attention). The examples which we have subjected to analysis have really not given us the right to take for granted a quantitative diminution of attention. We found what is probably not exactly the same thing, a disturbance of the attention through a strange obtruding thought. VII FORGETTING OF IMPRESSIONS AND RESOLUTIONS If any one should be inclined to overrate the state of our present knowledge of mental life, all that would be needed to force him to assume a modest attitude would be to remind him of the function of memory. No psychologic theory has yet been able to account for the connection between the fundamental phenomena of remembering and forgetting; indeed, even the complete analysis of that which one can actually observe has as yet scarcely been grasped. To-day forgetting has perhaps grown more puzzling than remembering, especially since we have learned from the study of dreams and pathologic states that even what for a long time we believed forgotten may suddenly return to consciousness.