HABIT FORMATION IN THE DOG 5 



experimenters' have pointed out, and as appears in the behavior 

 of every animal which I have used in discrimination tests, the 

 animals quickly form the habit of reacting in a certain rh\i;hm 

 or in a regular order or by taking a certain position, regardless 

 of the stimulus presented. These "position habits ' ' and rhythmic 

 choices are exceedingly hard to break up. If the order of pre- 

 sentation of the stimuli happens to coincide with or approx- 

 imate the habitual order adopted by the animal, a very high 

 record, or one even of perfect accuracy may be obtained w^hen 

 the animal really has not been affected by the stimulus.* And 

 indeed nothing is easier than for the experimenter to fall into a 

 rhythmic order of presenting the stimuli unless he determines 

 beforehand a chance order of presentation for each day's series 

 of trials. This question should be borne in mind while Kalischer's 

 results are being considered, for we shall have occasion to con- 

 sider it later. 



A further criticism may be made at this point. Kalischer 

 writes that the harmonium suited his purposes better than the 

 piano and some other instruments, because with the harmonium 

 it is possible to sustain the tone as long as is desired — or until 

 the animal has reacted. Now, it should not be forgotten that 

 the duration of the stimulus may easily become the basis of the 

 animal's choice. In such an experiment as Kalischer's, when 

 the dog is to react to one tone and is under penalty to inhibit 

 reaction to the others, reaction may not occur immediately fol- 

 lowing the stimulus. If the animal is timid reaction will prob- 

 abh^ be delayed. There is a strong temptation, which I have 

 experienced, to sustain the stimulus-tone a little longer if the 

 animal does not react properly or promptly, so as to give him 

 full opportunity to hear the tone. On the other hand the ex- 

 perimenter is likely to become content quicldy if the animal 

 does not react immediately to a "Gegenton" and damp the 

 stimulus ; whereas the animal might have reacted had the tone 

 been held a little longer. Rothmann, whose work we shall 

 presently discuss ^^ tells us that in his experiments the dura- 

 tion of the stimulus-tone was a disturbing factor, although he 



* c. f., Yerkes, R. M. The Dancing ]\Iouse, p. 111. 



* An actual record of one of my animals, nicely illustrating this point is shown 

 on page 52. 



1° RoTHMAXX, Max. Ueber die Ergebnisse der Horprufung an dressierten 

 Hunden. Arch. /. Physiol., 1908, pp. 103 ff. 



