APPENDIX 451 



edly believes in his own case. There are of course others about 

 the horses, grooms and attendants, to whom less attention has been 

 paid, but it is to be remarked that experiments have been made in 

 the absence of Herr Krall, and also in the absence of any groom, 

 so that apparently the horse is in contact with no one but the ex- 

 perimenter himself. Under such circumstances complicated sums 

 have been solved ; for example, Monsieur G. Bohn reports that in 

 the absence of Krall and the groom he has seen the following 

 performance : 



,4/282576? - V53H4I = H 



Close attention has been given to the behaviour of Krall and the 

 groom and nothing in the nature of a sign has been detected. To 

 this, however, there is one exception. In order to eliminate the 

 possibility of visual signs, Krall acquired a blind horse called 

 Berto. Herr Wigge, witnessing a performance with Berto, 

 observed that the groom twitched the bridle at the point when the 

 horse had stamped the right number. He had the two reins in 

 his hand and jerked the right or left for the units or the tens. 1 

 Herr KralPs reply to this accusation is that what Herr Wigge was 

 witnessing was not a demonstration of a performance but of the 

 method of teaching a beginner. But this reply is not wholly 

 satisfactory. Dr. von Maday 2 points out that no mention is made 

 of the use of the bridle in any of Krall's accounts of training ; that 

 the incident related by Wigge occurred after Berto had had six 

 weeks' training, whereas Krall maintains that Berto could count 

 up to nine in a fortnight. However, Haenel, one of KralFs most 

 vigorous supporters, six weeks later than Wigge, mentions the 

 substitution of a pull for the actual guidance of the feet, and says 

 that he was witness of the important moment at which the hand 

 could be left off. At the lowest, this indicates serious lacunas in 

 Krall's account of the training of the horses. It confirms the 

 suspicion, based above on internal evidence, that the horses do not, 

 at least at the outset, really count when Krall supposed them to 

 be doing so, but were responding to a signal, which he did in fact 

 use. But if this throws an element of doubt on the method or 

 training, it does not enable us to account for the actual results. 



To the theory that the horses give their replies in answer 

 to involuntary signs, there are four replies. The first is that, 

 with the one exception noted, the signs are incapable of 

 detection. This has already been considered. The second depends 

 on what we may call the loophole experiments. The third on 

 experiments in the absence of Krall and the groom ; and the 

 fourth on experiments in which the answer is wholly unknown 

 even to the experimenter himself (Unwissentliche Versuche). 



1 Tier seek ^ Heft I. p. 59. 2 Gibt es denkende Tiere, pp. 241-2. 



G G 2 



