2 G. V. HAMILTON 



But Yerkes' method, unlike mine, supplies a specifically direc- 

 tive feature which is deliberately introduced with the intention 

 of enabling the subject to form a definite, specifically adaptive 

 habit. The construction of his apparatus is such that now one, 

 now another group of apparently possible exit places (or their 

 equivalents) can be exposed to the subject, with trial to trial 

 variations of the position and constitution of such groups. For 

 example, in one series of trials by the Yerkes' method any three, 

 five, seven, nine or eleven of twelve exit places may be exposed 

 to the subject, who will invariably find that the middle exit 

 place of any three, five, seven, nine or eleven that may be ex- 

 posed is the right one to try. In other series it is the place at 

 the extreme left or at the extreme right of whatever group the 

 experimenter may expose that should be sought for escape. 



Watson's comparison of the two methods (5) implies a sim- 

 ilarity of intention which does not exist. While it is true that 

 Yerkes' method, like most methods that are employed for the 

 study of mammalian behavior, affords information as to various 

 types of searching reaction, it does so only incidentally, and 

 since it encourages habit formation rather than prolonged varia- 

 bility of response its intention is practically opposite to that 

 of my method. 



Watson's (5) reference to my method as a multiple stimuli 

 one, is in my opinion, somewhat misleading. One would not 

 refer to ordinary maze experiments as conforming to the multiple 

 stimuli type, since they employ confinement as the primary and 

 unvarying stimulus to reaction. Even where this stimulus is 

 reinforced by the introduction of features which render con- 

 finement particularly apt to lead to prompt reaction (e.g., elec- 

 trical stimulation within the enclosure) it remains, in a sense, 

 a single, general stimulus. The descriptions of my method and 

 apparatus that are given in subsequent pages will show that 

 unless the term " multiple stimuli method " may properly be 

 used to designate maze experiments, my method cannot be 

 regarded as a multiple stimuli one. 



The relation of my work to a general behavioristic program 

 cannot be satisfactorily defined without reference to the stand- 

 point from which I have proceeded to a study of behavior. 



A mammalian's initial adjustments to a situation which is 

 not characteristically encountered by the members of its species 



