BEHAVIOR OF VERTEBRATES 



433 



The following is one of the arrangements which Krall has 

 for use: 



DB 



G 



R 



B 



BB 



Here are some directions for the use of the diagram:— 



Multiply the two right hand numbers, r. 24. Add them, 

 r. 10. 



Multiply DB and BB. r. 42. Multiply DB and R. f. 42, 

 f. 34, r. 35. 



He sometimes writes the directions on the board as: "Adire 

 vangt troa + dus." Ans. f., f., r. 35. 



Phonetic spelling, which Krall says was acquired spontane- 

 ously, is strongly urged as a proof of intelligence. That this 

 varies as between horses is also argued as speaking against 

 training. For instance the spelling of the word horse varies 

 in Muhammed and ZarifT and each at different times have spelled 

 it as follows: — 



Muhammed — bfert, bfrt, fard, fart, fert, frt, faart, faerd. 

 faert, farb, fpferd, frrt, pard, pfart, ppverd, pfer, pferd, tfert, 

 fed, etc. 



ZarifT — bferd, fared, fferwt, fvert, pfrde, sdfert, pfert, bffet, 

 fdaerp. 



In the discussion of Claparede's articles before the French 

 society of Philosophy the arguments as to the intelligence of 

 these horses were clearly stated. Claparede was asked if he 

 thought the fact that the horses had different alphabets was 

 a proof of originality — a proof against the training hypothesis. 

 He was asked whether the obscurity in which the tests were 

 frequently made would not favor signs of some kind provided 

 there was any trickery. Some one else wanted to know how he 

 accounted for the tapping which frequently went on indefinitely. 

 Does the horse ever rectify a mistake was one question. Clapa- 

 rede, to this, replied that he had never seen a mistake rectified 

 by a horse unless there were signs on the part of some one 



