594 W. T. JAMES 



volving danger. From the present experiments, however, this 

 is doubtful. In the dogs of group B which responded to 

 every signal introduced, it was almost impossible to form a 

 negative response without the introduction of another reac- 

 tion with the negative signal which would involve lowering 

 the leg. This was attempted by presenting food with the 

 signal. While this might be considered an incentive, it is 

 really the involvement of another reflex system in the ex- 

 perimental situation, and an attempt to balance one against 

 the other in terms of the excitation-inhibition ratio. When 

 food is presented, the dog must shift the posture and approach 

 the pan, and this reaction makes the leg segment negative, 

 in contrast to its former positive movement to the signal. 

 Another plan involves the employment of the other foreleg. 

 If a second signal is accompanied by a shock on the left instead 

 of the right foreleg, this, of course, will elicit a positive 

 action of the left leg, and a negative action of the right leg. 

 By contrasting the two signals and shocks on each leg, the 

 dog may form a differentiation reaction. As a matter of 

 fact, it is this type of balancing one system against another 

 that involves the greater part of behavior, and for this 

 reason it should be considered in reflex experiments. Even 

 with such methods, however, "good" results were obtained 

 only among the mid-group animals. 



In the present investigation, no emphasis has been placed 

 on the limits of differentiating ability. The interest was in 

 the dog's natural tendency to react to every signal introduced 

 and in the response itself. Close differentiation is concerned 

 with cortical analysis; we were interested in how the wide 

 variety of visceral systems affect the total response rather 

 than in determining the analytical ability of each dog. 



UNTRAINABLE TYPES 



The experiments just described show the range of behavior, 

 or mode of adjustment, in two controlled situations. In this 

 connection it would be logical to ask if all dogs can be con- 

 ditioned to these situations and classified accordingly. Any- 



