cated a kind of recklessness in the observer's mind; a more disturbed condition than is suggested by the single occurrence of a significant association. The latter would happen to an observer who had no thought of concealment at all; the former, one might suppose, would be more likely to occur with one who had made up his mind that the best way to conceal would be to say whatever came into his head. However, the fact that such repetitions did occur with reference to an object that had not been seen at all, shows that it is unsafe to argue in this manner.
Over against these we may place the cases giving evidence of forced avoidance of significant associations. In eight instances a correct judgment was reached partly or wholly on the ground that not one of the relevant words suggested anything connected with the object to which it was intended to refer; that is, the avoidance of significant associations was complete. It was concluded that such avoidance must have been intentional. But there were two cases in which a wrong judgment was reached on this basis; not a single significant association occurred, and yet there was no intention to avoid, since the object to which the relevant words referred had not been seen at all. Then there were three correct judgments made on the basis of avoidance of a relevant association under circumstances that would naturally have suggested it: (1) after 'king' had suggested 'card,' 'diamond' suggested 'bracelet,' whereas it would in the mind of a person whose associations were normal have continued the 'card' train of ideas; (2) after 'tick' had suggested 'stop-watch,' 'second' suggested 'street;' (3) after 'fob' had suggested 'watch fob,' 'mainspring' suggested 'Main Street.' Here again we have to offset these instances with a case of failure by the same method, but the mistake occurred where the application of the principle is by no means so clear as it was in the series just mentioned. After 'spade' suggested 'hearts,' 'heart' suggested 'feelings;' it was wrongly concluded that the pack of cards had been seen, but the passage to a different line of associations here might have been due to the fact that the stimulus word 'heart' was a repetition of the last reaction word, so that the observer was not unnaturally given an impulse to break away from that set of ideas in a new direction. The effort to avoid significant associations shows itself also, as has often been pointed out, in the occurrence of far-fetched associations. The following is a list of the associations of this character upon which correct judgments were based: 'calico—variegated colors,' 'fang—boy,' 'second—how many in an hour,' 'time—big idea,' 'deal—what kind,' 'Elgin—make,' 'mainspring—mechanism,' 'tick—glasses,' 'second—tie,' 'time—guess,' 'slimy—dessert,' 'heart—chicken,' 'Elgin—Europe,' 'mainspring—lake,' 'tick—year,' 'mainspring—time,' 'hour—National Park.' In some series two or more of these 'forced' associations were found; altogether, eleven correct judgments were based wholly or partly upon them. They may not, as reported above, seem particularly far-fetched in every case, but they were always 'peculiar,' as compared with the other associations in the series. Three erroneous conclusions were formed on account of the occurrence of associations, at least as forced as any of those above, viz.: 'fang—fishes,' 'deal—page,' 'card—hardness' (the last two occurred in the same series), 'wriggle—little girl of that name.' Again, the effort to avoid is betrayed by the occurrence of sound or rhyme associations. These were found in the case of two observers only, and a correct conclusion was drawn in each case. One gave the following associations: 'tick—click,' 'time—last time,' 'hour—glower,' 'gold—golden;' the object seen being the gold watch. The other observer, who had seen the snake, gave:
