The largeness of the number 33 in the middle entry of the last column but one, which disconcerts the run of the series, is wholly due to a visual memory of places seen in manhood.  I will not speak about this now, as I shall have to refer to it farther on.  Neglecting, for the moment, this unique class of occurrences, it will be seen that one-half of the associations date from the period of life before leaving college; and it may easily be imagined that many of these refer to common events in an English education.  Nay further, on looking through the list of all the associations it was easy to see how they are pervaded by purely English ideas, and especially such as are prevalent in that stratum of English society in which I was born and bred, and have subsequently lived.  In illustration of this, I may mention an anecdote of a matter which greatly impressed me at the time.  I was staying in a country house with a very pleasant party of young and old, including persons whose education and versatility were certainly not below the social average.  One evening we played at a round game, which consisted in each of us drawing as absurd a scrawl as he or she could, representing some historical event; the pictures were then shuffled and passed successively from hand to hand, every one writing down independently their interpretation of the picture, as to what the historical event was that the artist intended to depict by the scrawl.  I was astonished at the sameness of our ideas.  Cases like Canute and the waves, the Babes in the Tower, and the like, were drawn by two and even three persons at the same time, quite independently of one another, showing how narrowly we are bound by the fetters of our early education.  If the figures in the above Table may be accepted as fairly correct for the world generally, it shows, still in a measurable degree, the large effect of early education in fixing our associations.  It will of course be understood that I make no absurd profession of being able by these very few experiments to lay down statistical constants of universal application, but that my principal object is to show that a large class of mental phenomena, that have hitherto been too vague to lay hold of, admit of being caught by the firm grip of genuine statistical inquiry.  The results that I have thus far given are hotch-pot results.  It is necessary to sort the materials somewhat before saying more about them.

After several trials I found that the associated ideas admitted of being divided into three main groups.  First there is the imagined sound of words, as in verbal quotations or names of persons.  This was frequently a mere parrot-like memory which acted instantaneously and in a meaningless way, just as a machine might act.  In the next group there was every other kind of sense imagery; the chime of imagined bells, the shiver of remembered cold, the scent of some particular locality, and, much more frequently than all the rest put together,