by association, instantly brings the whole distinctly before him. There are even facts which seem to show that the impression recalled by local associa, tion may affect the bodily organs. Van Swieten relates of himself, that he was passing a spot where the dead body of a dog burst and produced such a stench as made him vomit; and that, happening to pass the same spot some years after, he was affected by sickness and vomiting from the recollection.

Finally, to the influence of local association we are to refer the impressions produced by the monuments of the illustrious dead; the trophies of other times; the remains of Greece and Rome; or by the visitation of spots distinguished by illustrious deeds, as Thermopylæ, Bannockburn, or Waterloo, "Far

from me," says Dr. Johnson, "and from my friends, be such frigid philosophy, as may conduct us, indifferent and unmoved, over any ground which has been dignified by wisdom, bravery, or virtue. That man is little to be envied whose patriotism would not gain force upon the plains of Marathon, or whose piety would not grow warmer among the ruins of Iona."

III. ARBITRARY or FICTITIOUS ASSOCIATION.-This association is generally produced by a voluntary effort of the mind; and the facts associated are not connected by any relation except what arises out of this effort. The process is exemplified in the connexion we establish between something which we wish to remember and something which we are in no danger of forgetting; as in the common expedients of tying a thread about the finger, or making a knot on the pocket-handkerchief. A Roman, for the same purpose, turned the stone of his ring inwards towards the palm of his hand. There is an analogous expedient which most people probably have employed for enabling them to remember the names of persons. It consists in forming an association between the name to be remembered and

that of some intimate friend or public character of the same name, which is familiar to us. The remarkable circumstance in these cases is, that whatever difficulty a person may have in simply remembering a name, he never forgets who the individual was with whose name he formed the association.

On this principle have been founded various schemes of artificial memory. One of the most ancient consisted in associating the divisions of a discourse to be delivered with the various apartments of a building, and the leading sentiments with articles of furniture. This is said to have been much practised by the ancient orators, and to have given rise to the phraseology by which we speak of the divisions of a discourse, as the first place, the second place, &c. I have repeatedly made experiments on this method in remembering the discourses of public speakers, and the effect is certainly astonishing; for though it is many years since the experiments were made, I still find articles of furniture associated in the clearest manner with sentiments delivered by some of the speakers. Other systems of artificial memory are founded upon the same general princi ple, though the particular applications of it may vary; and some of them are extremely absurd. One of the last which attracted notice in this country was that of a German of the name of Feinagle, who de livered lectures on memory to crowded and fashion able audiences, about the year 1809 or 1810. A leading part of his system was the memory of dates, and it consisted in changing the figures in the date into the letters of the alphabet corresponding to them in number. These letters were then formed into a word to be in some way associated with the date to be remembered. One example, which I happen to recollect, will be sufficient to illustrate the peculiarity of the system, and at the same time its efficiency for its purpose. Henry IV. King of England was born in the year 1366. This date, changed into letters,

66

gives mff, which are very easily formed into the word muff. The method is not so obvious of establishing with this a relation to Henry IV. Henry IV.," says M. Feinagle, "is four hens, and we put them into the muff, one in each corner.' "No one, certainly, after hearing this, is in any danger of forgetting the date of the birth of Henry IV.; but whether the remembrance is worth such a process is a separate question.

It is unnecessary to enlarge upon the subject of arbitrary association, as the observation of every one will furnish numerous examples of it. There is one application of the principle, however, which deserves to be referred to in a more particular manner. I allude to the practice of commemorative rites, or periodical observances, for transmitting the remembrance of remarkable events. These are in their nature, in general, entirely arbitrary; or, if they have any analogy to the events, the relation is only figurative. But the influence of such celebrations is of the most extensive and most important kind. If the events, particularly, are of a very uncommon character, these rites remove any feeling of uncertainty which attaches to traditional testimony, when it has been transmitted through a long period of time, and consequently through a great number of individuals. They carry us back, in one unbroken series, to the period of the events themselves, and to the individuals who were witnesses of them.

The most important application of the principle in the manner now referred to is in the observances of religion which are intended to commemorate those events which are connected with the revelation of the Christian faith. The importance of this mode of transmission has not been sufficiently attended to by those who have urged the insufficiency of human testimony to establish the truth of events which are at variance with the common course of nature. We have formerly alluded to one part of this sophism,

and have stated the grounds on which we contend that no objection to the credibility of these events can be founded upon our observation of what we call the course of nature. We have admitted that a much higher species of evidence is required for them than would be required for events which correspond with our previous observation; and this high and peculiar evidence is confirmed in a striking manner by the periodical rites now referred to. By means of these we are freed entirely from every impression of the fallibility of testimony, and the possibility of the statements having been fabricated; as we are conducted in one uninterrupted series to the period when the events took place, and to the individuals who witnessed them. This will appear if we state in a few words a hypothetical case. Let us conceive a person attempting to impose upon the world by an account of some wonderful or miraculous event, which he alleges occurred five hundred years ago. He, of course, exerts every possible ingenuity in fabricating documents, and framing the appearance of a chain of testimony in support of his statement. It is quite possible that he might thus deceive a considerable number of credulous persons; and that others, who did not believe his statement, might yet find difficulty in proving its fallacy. But if the report were further to bear, that ever since the occurrence of the alleged event it had been regularly and specially celebrated by a certain periodical observance, it is clear that this would bring the statement to the test of a fact open to examination, and that the fallacy of the whole would be instantly detected.

On these principles it must appear that the statements of the sacred writings, respecting miraculous events which are said to have occurred upwards of 1800 years ago, could not have been fabricated at any intermediate era during that period. It is unnecessary to state how much more improbable it is that they could have been fabricated at the very time and K

place in which they are said to have occurred, and in the midst of thousands who are said to have witnessed them, many of whom were deeply interested in detecting their fallacy. This part of the question is not connected with our present inquiry, but it is impossible to dismiss the subject without one reflection:-that if we are to proceed upon the principle of probabilities, we must balance fairly the probabilities of fabrication. If we do so, we hesitate not to assert, that the probability of the world being imposed upon, under all the circumstances now alluded to, is more at variance with our firm and unalterable experience than all that we are called upon to believe.

It does not appear necessary to say much of that modification of memory which is called CONCEPTION. It is the recalling of a perception. If, for example, we have passed a person in the street whose face we think we have seen, but without being able to recognise him, we can recall the impression of his countenance, and endeavour to recollect who he is. By a higher exercise of this faculty a painter can draw from conception a landscape or a building long after he has visited them, and even the portrait of a friend who is dead or absent, and whom he has not seen for a considerable time. By another modification of this power we can imbody into a conception a scene, a figure, or a transaction which has been described to us by another. The vividness of our conception, in such cases, does not depend upon the accuracy or even the truth of the description, but upon the degree of liveliness with which it is given, or the intensity with which our attention is directed to it. Thus, it has been remarked that we have a more clear conception of Don Quixote or Sancho than of any characters in real history, unless they have been made familiar to us by paintings. The business of the novelist being to create his hero, he gives a more full and graphic delineation of him than the authentic