Dugald Stewart has endeavoured to account for these phenomena by the doctrine that in sleep the operations of the mind are suspended, and that therefore the cause of dreams is the loss of power of the will over the mind, which in the waking condition is subject to its control. Now, if this be the case, dreams must consist of mental operations independent of the will. However, it is not the suspension of the will and of the powers of volition that alone constitutes sleep; it is the suspension of the powers of the understanding,--attention, comparison, memory, and judgment. It is in consequence of this suspension of all our active intellectual faculties that we never can _will_ during our dreams; in that state there appears to be a resistance of the powers of volition with which the mind struggles in vain, and which is expressed both by moans, and the character of the sleeper's every feature, which portrays a state of anguish and impatience. In all dreams that are not of a morbid nature, every action is passive, involuntary. This state is widely different from delirium, in which the brain is in a morbid state of excitement; and the body is more susceptible than usual of external agency, while the mind is perplexed by hallucinations of an erroneous nature. Dr. Abercrombie considers insanity and dreaming as having a remarkable affinity when considered as mental phenomena; the impressions in the one case being more or less permanent, and transient in the other. Somnambulism he considers an intermediate state. Dreams, according to his theory, are divided into four classes: the first, when recent events and recent mental emotions are mixed up with each other, and with old events, by some feeling common to both; the second class relates to trains of images brought up by association with bodily sensations; the third, the result of forgotten associations; and the fourth class of dreams contains those in which a strong propensity of character, or a strong mental emotion, is imbodied in a dream, and by some natural coincidence is fulfilled. The following interesting cases that fell under Dr. Abercrombie's immediate notice, illustrate his views and the above classification. Regarding the first class, Dr. A. relates the following: "A woman, who was a patient in the clinical ward of the infirmary of Edinburgh, under the care of Dr. Duncan, talked a great deal in her sleep, and made numerous and very distinct allusions to the cases of other sick persons. These allusions did not apply to any patients who were in the ward at the time; but, after some observation, they were found to refer correctly to the cases of individuals who were there when this woman was a patient in the ward two years before." The following is an instance of phantasms being produced by our associations with bodily sensations, and tends to show how alive our faculties continue during sleep to the slightest impressions: