Xov. 25, 1881.] 



KNOWLEDGE 



69 



narrow limits of weiglit, he will obtain a much less satis- 

 factory result than he would if he used a common wcifjliing 

 machine incapable of showing differences of a few grains or 

 scruples. Yet he should not condemn the balance, but 

 himself for misusing it. 



Not unfrequently mathematicians haNe fallen into a 

 parallel error. Indeed, there are those who assert that 

 mathematicians are more apt than other men to err in this 

 way, because they try to apply the exact laws of proba- 

 bility to matters about which they only have inexact 

 knowledge. To use an illustration of Huxley's, mathe- 

 maticians possess a mill of exquisite perfection, which 

 grinds you stuff of any degree of fineness, but they are apt 

 to put peacods into it and to regard the product. of their 

 grinding as wheaten flour. 



However, there is no real connection between profound 

 mathematical research and utter absence of common sense. 

 Mathematicians do not often err in this way. The laws of 

 probability would be worth knowing even if they did. Nor 

 is the application of these laws to many cases where they 

 really are applicable likely to detract from what is com- 

 monly called the common sense way of viewing matters. 

 Indeed, the common sense method may be regarded as only 

 a rough way of applying the just method. The probability 

 of its leading to a tolerably correct result is much less 

 than the probability of a correct result when due 

 account is taken of details; just as the chance of 

 coiTectly estimating the cost of a complicated work by 

 roughly adding a number of roughly-estimated items is less 

 than the chance of a correct estimate when these items, 

 separately estimated as exactly as possible, are properly 

 added together. In neither case do we get an exact result, 

 but the chances of gross error are much greater in the 

 former than in the latter ; for to the chances of error 

 arising from an exact estimate of the several items in the 

 latter case there is added the chance of a large error in a 

 rough estimate of the totality of these items. 



It must be admitted that liy justly combining the \arious 

 probabilities on which a question depends, we are more 

 likely to attain a correct opinion than if we formed only a 

 rough idea of what would result from such a combination. 

 Yet the incautious application of the laws of probability, 

 combined with undue reliance on results which, at the 

 most, had been only shown to be probably true, has often 

 led to absurd blunders. 



INTELLIGENCE IN ANIMALS. 



PASSING over several cases whichseem to carry the matter 

 no further than those already cited, we come next to a 

 case which appearstous oneof the most .strikingever recorded. 

 Tlie wn-iter, Mr. E. H. Pringle, remarks that it is an instance 

 of sagacity which finally set at rest anj' doubts he had ever 

 entertained tliat the difference between human and animal 

 intelligence is one of degree only. We can see no way in 

 which the story can be explained without assuming the 

 exercise of something more than that mere practical reason- 

 ing which probably underlies all the so-called instinctive 

 actions of animals : — 



Mr. J. W. Chen-y, of the Madras Forest Service, is 

 owner of a dog, a bull-terrier, called " Bully." (This breed 

 is notoriously not the most intelligent of the canine tribe, 

 so that the behaviour of Bully appears all the more strik- 

 ing.) " We lived," says Mr. Pringle, " in the bungalow A* 

 the compound of which was bounded south and west by 



* The figure presents all tlie features essential to a correct under- 

 standinff of Jlr. Priiifrle's narrative. 



public roads DC and GFC, both leading to the cantonment 

 of Ijangalore in the direction C. There were three gates 

 into the compound at C, D, and G, the main appi'oach to 

 the bungalow leading over a bridge B that spaimed 

 a public road FD. The compound was tilled with 

 trees and shrubs, and bordered by dense lantana hedges, 

 so that with the exception of a portion of th" 

 western road at F, neither of the cantonment roads 

 were visible from the bridge, nor could the footpaths 

 be seen thence. Now, Bully had a lady friend (canine) 

 living in the cantonment, and at times she was so attractive, 

 that absences without leave on the part of the dog were 

 frequent. After one of these excursions. Bully had been 

 brought back and chained up for the night. Next morning, 

 wliile his master and I were sitting at early breakfast, it 

 was decided that he should be released, and to effectually 



stop further delinquency, a peon was sent down to the 

 bridge with orders to intercept him if he started for the 

 cantonment. Bully was brought in and unchained ; he had 

 that unmistakable air of detected guilt deservedly punished, 

 and spent some time in begging for scraps from the table in 

 a most deprecatory manner. Shortly, however, he strolled 

 into the \'erandah, and then down the front steps on to the 

 gravel walk. After wandering about aimlessly for a few 

 minutes, he quietly started down the approach AHB. We 

 followed, keeping out of his sight. At the turn of the road. 

 Bully met with the unexpected apparition of the peon 

 standing on the bridge. In a moment, though not a word 

 was spoken by the man, the dog turned and came straight 

 back to the room, whither we had in the meantime slipped 

 back unobserved, and re-entered it, wagging his tail vio- 

 lently, and looking exceedingly sheepish. He now lay down, 

 and closed his eyes. The cocked ears showed that sleep was 

 mere pretence, and he soon rose again, went out into the 

 front garden, and hunted for buried bones — purely imaginary 

 ones, I believe. His search gi'adually led him down the 

 hill by a footpath leading from A to G, we keeping him in 

 sight, as before, and he finally reached the road at the 

 bottom. There all disguise was dropped, and he started oft 

 for the cantonment. As he neared the spot F, the peon 



