distinguish and remember for a sufficient time the very impression or association which gave birth to each thought; but we can do this as far as it can be expected to be done, and in so great a variety of instances, that our argument for the prevalence of the foregoing principle of association in all instances, except those of new impressions, may be esteemed a complete induction.
A reverie differs from imagination only in that the person being more attentive to his own thoughts, and less disturbed by foreign objects, more of his ideas are deducible from association, and fewer from new impressions.
It is to be observed, however, that in all the cases of imagination and reverie, the thoughts depend, in part, upon the then state of body or mind. A pleasurable or painful state of the stomach or brain, joy or grief, will make all the thoughts warp their own way, little or much. But this exception is as agreeable to the foregoing theory, as the general prevalence of association just laid down.
We come next to dreams. I say then, that dreams are nothing but the imaginations, fancies, or reveries of a sleeping man; and that they are deducible from the three following causes; viz. First, The impressions and ideas lately received, and particularly those of the preceding day. Secondly, The state of the body, particularly of the stomach and brain. And, thirdly, Association.
That dreams are, in part, deducible from the impression and ideas of the preceding day, appears from the frequent recurrence of these in greater or lesser clusters, and especially of the visible ones, in our dreams. We sometimes take in ideas of longer date, in part, on account of their recency: however, in general, ideas that have not affected the mind for some days, recur in dreams only from the second or third cause here assigned.
That the state of the body affects our dreams, is evident from the dreams of sick persons, and of those who labour under indigestions, spasms, and flatulencies.
Lastly, We may perceive ourselves to be carried on from one thing to another in our dreams partly by association.
It is also highly agreeable to the foregoing theory to expect, that each of the three foregoing causes should have an influence upon the trains of ideas that are presented in dreams.
Let us now see how we can solve the most usual phænomena of dreams upon these principles.
First, then, The scenes which present themselves are taken to be real. We do not consider them as the work of the fancy; but suppose ourselves present, and actually seeing and hearing what passes. Now this happens, First, Because we have no other reality to oppose to the ideas which offer themselves, whereas in the common fictions of the fancy, while we are awake, there is always a set of real external objects striking some of our senses, and precluding a like mistake there: or, if we become quite inattentive to external objects, the reverie does so far put
