THE ANALYSIS OF BEHAVIOUR IN ANIMALS 167 



indifferent stimulus such as lighting a lamp before the dog's 

 eyes, the sound immediately regains its efficacy. The sound 

 and the light were each indifferent stimuli ; their combined 

 effect depends on the fact that the former had previously 

 been a conditioned stimulus, i.e. that the latter breaks down 

 the inhibition to which the former was temporarily subject. 

 The possible bearing of this phenomenon on the phenomenon 

 of " attention " is evident. How unexpected fields may be 

 illuminated by study of the conditioned reflex is well seen in 

 the phenomenon of experimental neurasthenia. When a new 

 internal inhibition is in process of formation a preformed 

 inhibition is weakened. Suppose a spheroidal object is 

 established as a conditioned stimulus for salivary secretion. 

 When an ellipsoid differing only in the length of one axis 

 is presented, it is at first an effective stimulus and ceases to 

 be so by differential inhibition as already described. Now 

 suppose that we successively present ellipsoids approaching 

 more nearly the spheroidal form, pushing the process to the 

 limit of discrimination, marked changes in the dog's behaviour 

 occur, firmly established inhibitions disappear, its excitability 

 is greatly increased. After two months' rest the previous 

 state is regained and old conditioned reflexes reappear. 



Finally, the study of conditioned reflexes as implied in 

 the last type of experiment opens up a new horizon for the 

 objective and quantitative treatment of sensation, aside from 

 the consideration of the sense-organs as physical apparatus. 

 This may be illustrated by employing the conditioned reflex 

 to define the limits of discrimination. The sound of a tuning- 

 fork of 256 vibrations (middle C) is accompanied by electrical 

 stimulation of the paw until it is established as a conditioned 

 stimulus. A tuning-fork of 264 vibrations presented as a 

 signal for withdrawal of the paw evokes response which 

 subsides on successive presentation before the effect of the 

 original conditioned stimulus is extinguished by internal 

 inhibition, as may be tested by applying it. A series of pairs 

 of forks with diminishing differences in tone are now tried 

 out till no differential inhibition can be established for a 

 given pair. The limit of discrimination for sound in dogs is 



