446 JOHN LINCK ULRICH 



limbs. In addition, directive integration can not be said to be 

 facilitated by an association of the senses, when its facilitation 

 is shown to be primarily dependent upon the functional condi- 

 tion of the entire organism. " Sensory association," if established 

 in learning, should, in a conclusive way, modify the nervous 

 system and remove the fluctuations of the extensor thrusts 

 which produce imperfect responses, disrupting directive integra- 

 tion from time to time. The extensor thrusts play an exceed- 

 ingly important part in learning. All the movements made by 

 a rat seem to be associated with the fundamental movement of 

 extension and flexion of the limbs in progression, and if the 

 nervous system is modified in a definite way in learning, this 

 movement ought first to be developed in each limb and the 

 unequal antagonistic action of the limbs disappear, before any 

 changes take place in the nervous system. Fluctuations and 

 the undeveloped condition of the extensor thrusts persist. Sen- 

 sory excitations do not, it seems, predominantly function in 

 learning so as to modify in any definite way the nervous system; 

 that is, there is no specific " sensory control" of movements as is 

 implied in all present theories of learning. 



With no direct evidence of a definite modification in the 

 nervous system taking place in learning, it is difficult to assume 

 that the reappearance of specific sensory excitations produce in 

 succeeding trials a fixation or a retention of a " successful" 

 movement. With an effective adjustment to every part of the 

 environment outside of the problem box and an exaggerated 

 performance of progression, and the neck reflex movement for a 

 greater adjustment to .the problem box to raise the latch, it 

 is hardly necessary to posit any additional view that new move- 

 ments are fixated. The exaggerated movements persist, because 

 they are, of all the movements made, the most effective in an 

 adjustment to a particular configurated part of the problem box, 

 when previously other parts of the box have been effectively 

 adjusted to. The occasional use of the teeth to raise the latch is 

 another method but it is much less effective. Another con- 

 sideration, is that progression and the neck reflex movement are 

 inherently conjoined, and it is not necessary to assume that a 



