REASONING. 351 



dogs carrying coppers to pastry-cooks to get buns, and it is 

 said that a certain dog, if lie gave two coppers, would never 



e.g. collies, seem instinctively to defend their master's property. The case 

 is similar to that of a dog's barking at people after dark, at whom he would 

 not bark in daylight. I have heard this quoted as evidence of the dog's 

 reasoning power. It is only, as Chapter III has shown us, the impulsive 

 result of a summation of stimuli, and has uo connection with reasoning. 



In certain stages of the hypnotic trance the subject seems to lapse into 

 the non-analytic state. If a sheet of ruled foolscap paper, or a paper with 

 a fine monotonous ornamental pattern printed on it, be shown to the sub- 

 ject, and one of the ruled lines or elements of the pattern be pointed to for 

 an instant, and the paper immediately removed, he will then almost always, 

 when after a short interval the paper is presented to him again, pick out the 

 indicated line or element with infallible correctness. The operator, mean- 

 while, has either to keep his eye fixed upon it, or to make sure of its posi- 

 tion by counting, in order not to lose its place. Just so we may remember 

 a friend's house in a street by the single character of its number rather 

 than by its general look. The trance-subject would seem, in these instan- 

 ces, to surrender himself to the general look. He disperses his attention 

 impartially over the sheet. The place of the particular lino touched is part 

 of a ' total effect ' which he gets in its entiretjs and which would be distort- 

 ed if another line were touched instead. This total effect is lost upon the 

 normal looker-ou, bent as he is on concentration, analysis, and emphasis. 

 "What wonder, then, that, under these experimental conditions, the trance- 

 subject excels hira in touching the right line again ? If he has time given 

 him to count the line, he will excel the trance-subject ; but if the time be too 

 short to count, he will best succeed by following the trance-methc i, ab- 

 staining from analysis, and being guided by the ' general look ' of the line's 

 place on the sheet. One is surprised at one's success in this the moment one 

 gives up one's habitually analytic state of mind. 



Is it too much to say that we have in this dispersion of the attention 

 and subjection to the ' general effect ' something like a relapse into the 

 state of mind of brutes? The trance-subject never gives any other reason 

 for his optical discriminations, save that ' it looks so.' So a man, on a road 

 once traversed inattentively before, takes a certain turn for no reason ex- 

 cept that h.& feels as if it must be right. He is guided by a sum of impres- 

 sions, not one of which is emphatic or distinguished from the rest, not one 

 of which is essential, not one of which is conceived, but all of which 

 together drive him to a conclusion to which nothing but that sum-total 

 leads. Are not some of the wonderful discriminations of animals expli 

 cable in the same way ? The cow finds her own stanchions in the long 

 stable, the horse stops at the house he has once stopped at in the monoto- 

 nous street, because no other stanchions, no other house, yield impartially all 

 the impressions of the previous experience. The man, however, by seek- 

 ing to make some one impression characteristic and essential, prevents the 

 rest from having their effect. So that, if the (for him) essential feature be 

 forgotten or changed, he is too apt to be thrown off altogether, and then 

 the brute or the trance-subject may seem to outstrip him in sagacity. 



Dr. Romanes's already quoted distinction between ' receptual ' and 



