62 KNIGHT DUNLAP 



tion of the act, or by the dwelling in thought on the act, which 

 is probably physiologically equivalent to a repetition. These 

 cases come under the heading of repetition or frequency, which, 

 so far as I know, no one doubts to be an important factor in 

 association or habit-formation. The cases which cannot be 

 subsumed under this head are the ones in which the efficacy of 

 pleasure has been challenged, and these are the ones with which 

 we are at the present moment concerned. 



If pleasure (to neglect pain for the moment) is directly con- 

 nected with a change in internal secretions, and if internal secre- 

 tions may act on the nervous mechanism (both of which condi- 

 tions are possible), we have in this aspect of pleasure a possible 

 means of influencing habit formation. Since the reaction which 

 we suppose to be " fixed 7 ' by pleasure precedes rather than fol- 

 lows the pleasure; or at least precedes the hedonic secretory 

 effects as we must allow a time interval for the secretion to be 

 carried in the blood stream to the effective locality the influ- 

 ence of whatever hormone is involved is retroactive, i.e., it will 

 act on a pathway over which discharge has occurred in such a 

 way as to make discharge over that pathway more ^probable 

 in the future than it was before. 



In brief, the obviously suggested theory is that the nervous 

 discharge leaves an arc or certain important pbints in the arc in 

 such a condition chemically, that a certain substance (hormone) 

 may a few moments later "fix" it. Artificial as this theory 

 sounds at first, I believe it is worth putting to the test. 



I might point out familiar observation, and data from the 

 experimental investigations of learning, which fit this working 

 hypothesis. There are many pertinent cases. I believe how- 

 ever that this procedure would not be legitimate, since the ob- 

 servations which support my hypothesis were not made with 

 this hypothesis in view, and hence the strong backing they afford 

 may be more apparent than real. The important thing is that 

 experimenters should hereafter keep the hypothesis in mind, and 

 observe specifically in future work the data which have direct 

 bearing on it. 



The particular efficacious homone which is liberated in pleasure 



