432 JOHN LINCK ULRICH 



dorsally extended. On the other hand, the records from the 

 second group show that the reflex thrusts fluctuated markedly 

 and often. At times nearly all the reflex thrusts could be 

 evoked. In addition, the first rat of this group showed evident 

 fluctuations in reflex excitability. Fluctuations in excitability 

 have frequently been observed, and then accompanied at times 

 with marked fluctuations in reflex extensor thrusts. To such an 

 extent did fluctuations occur in rats whose record is presented in 

 table 8, that the reaction time for directive integration and the 

 neck reflex movements could not at times be obtained. Wander- 

 ing progression from the entrance box occurred instead of direct 

 progression to the door, and the reaction times were impossible 

 to obtain by one revolution of the kymograph. Consequently, 

 blank spaces occur in the table when reflex excitability fluctuated. 

 Frequently great increases in the reaction times for directive in- 

 tegration and for the neck reflex appear. With the second rat 

 of this group instead of progression being direct to the latch, 

 it was often around the problem before the latch was raised. 

 Again the reaction time for directive integration and for the neck 

 reflex could not be obtained. Though reflex excitability was 

 manifest in this rat, there was great difficulty in posturing under 

 the latch to produce the neck reflex movement, and this appears 

 to be due to active and frequent fluctuations of the reflex ex- 

 tensor thrusts. 



The results from the rats presented in tables 8 and 9 are 

 different from those obtained from any rats experimented with, 

 and this shows that as far as these experiments go, they do not 

 exhaust what may still occur when learning the latch-box prob- 

 lem, particularly is this so when reflex excitability can not 

 readily be judged to be hyponormal. The individual functional 

 condition of rats varies greatly. The record of the first of these 

 two rats shows that though reflex excitability was hypernormal 

 in the first trial, activity was constant but slow. The problem 

 was solved on the first day, but in later trials reflex excitability 

 became after a marked fluctuation hyponormal. The record of 

 the second rat shows that reflex excitability was always mani- 

 fest, the problem was solved on the first day, but learning was 



