6 HARRY MILES JOHNSON 



reports no precautions against it, and does not consider it in 

 interpreting his data. He says: "After three weeks (of train- 

 ing) the dogs reacted with perfect accuracy; only now and 

 then, at a tone of abnormally long duration, a false reaction 

 occurred." " Again, speaking of animals bereft of one posterior 

 corpus quadrigeminum, which were hardly disturbed by the 

 operation : "If the tones were sounded for a somewhat longer 

 time, then the dogs came at every tone." '' To guard against 

 disturbance from this source, care should be taken to make the 

 duration of all stimuli the same at every presentation. Before 

 Kalischer and Rothmann assumed that their animals were dis- 

 criminating only the pitch of tones, they ought to have shown 

 in control tests that their animals would not react to a " Gegen- 

 ton" no matter how long was its duration.^ =* 



Kalischer does not tell us how many trials he gave the animals 

 at each daily test, nor how many trials altogether were required 

 to achieve perfection. Data on both of these questions are 

 highly desirable in reports of behavior experiments. He does 

 say in this report, however, that the daily test on each animal 

 lasted not longer than five or six minutes. In reporting another 

 experiment of similar character he tells us that about fifteen 

 minutes were required to give eight to ten trials'*. This 

 tallies with the writer's experience. An animal should not be 

 worked much faster than this, as the risk of its becoming "in- 

 attentive" is too great. But it is evident that the stimulus 

 cannot be presented over three to five times in the five or six 

 minutes which Kalischer allowed for a daily series. The possi- 

 bility of varying the order of presentation of food-tone and 

 "Gegentone" is also greatly limited by the brevity of the series. 

 For this reason it is doubtful whether the animal is given a 

 fair test of discrimination. An animal making random choices 

 only may chance to make most of the first three or four of a 

 series correctly, and yet be unable to maintain accuracy through 

 a varied series of ten to twenty trials. On the other hand, some 



" Loc. cit., p. 107. 



!=> Loc. cit., p. 109. 



1^ It may be said here that it is quite possible to make tests of the clog's sensi- 

 tivity to mere duration-ditference of auditory stimuli. The results of such an 

 investigation should prove very interesting. 



" Kalischer, Otto. Weitere Mitteilung ueber die Ergebnisse der Dressur 

 als physiologischer Untersuchungsmethode auf den Gebieten des Gehor — Geruchs- 

 und Farbensinns. Arch. f. Physiol., 1909. 



