tributes to fix them, and their order, in the memory, in the same manner as the frequent impression of the objects themselves.
Thirdly, Suppose now a person so far advanced in life, as that he has learnt all these rudiments, i.e. that he has ideas of the common appearances and occurrences of life, under a considerable variety of subordinate circumstances, which recur to his imagination from the slightest causes, and with the most perfect facility; and let us ask, how he can be able to remember or recollect a past fact, consisting of one thousand single particulars, or of one hundred such clusters as are called the rudiments of memory; ten single particulars being supposed to constitute a rudiment? First, then, We may observe, that there are only one hundred links wanting in the chain; for he has already learnt considerable exactness in the subordinate circumstances of the one hundred clusters; and perfect exactness is not to be supposed or required.—Secondly, The one hundred clusters recur again and again to the imagination for some time after the fact, in a quick and transient manner, as those who attend sufficiently to what passes in their own minds may perceive; and this both makes the impression a little deeper, and also serves to preserve the order. If the person attempts to recollect soon after the impression, the effect remaining in the brain is sufficient to enable him to do this with the accuracy required and experienced; if a longer time intervene, before he attempts to recollect, still the number of involuntary recurrences makes up in some measure for the want of this voluntary recollection. However, the power of recollection declines in general, and is entirely lost by degrees. It confirms this reasoning, that a new set of strong impressions destroys this power of recollection. For this must both obliterate the effects of the foregoing impressions, and prevent the recurrency of the ideas.—Thirdly, As the single impressions, which make the small clusters, are not combined together at hazard, but according to a general tenor in nature, so the clusters which make facts succeed each other according to some general tenor likewise. Now this both lessens the number of varieties, and shews that the association between many of the clusters, or rudiments, or one hundred links supposed to be wanting, is cemented already. This may be both illustrated and exemplified by the observation, that it is difficult to remember even well-known words that have no connexion with each other, and more so to remember collections of barbarous terms; whereas adepts in any science remember the things of that science with a surprising exactness and facility.—Fourthly, Some clusters are excluded from succeeding others, by ideas of inconsistency, impossibility, and by the methods of reasoning, of which we become masters as we advance in life.—Fifthly, The visible impressions which concur in the past fact, by being vivid, and preserving the order of place, often contribute greatly to preserve the order of time, and to suggest the clusters which may be
