584 



BRAIN MECHANISMS AND LEARNING 



electrophysiological experiments (Adcy, Sunderland and Dunlop, 1957; 

 Adcy, Dunlop and Sunderland, 1958). In the trained animal, the phase 

 ciistribution is totally reversed, taking on an aspect consistent with path- 

 ways suggested by Cajal, running from the entorhinal area to the dorsal 

 hippocampus. 



The probable significance ol these fnidings is turther exemplified by the 

 observations that in animals in late training, awaiting the opportunity to 

 make an approach periormance, auto-corrclations showed considerable 



CROSS-CORRELATIONS — IN EARLY TRAINING 

 CAT 7 — TRIAL 4— RUN 14 



40 SAMPLINGS/ SEC. BANDPASS I-50C.P.S 



A. 



DORSAL HIPR LEADS q-S- ENTORHINAL AREA LEADS 



(REGION CA4> _»i„_35MSEC 



"t 



B. 



TRIAL 4 - RUN 18 



. DORSAL HIPP LEADS 

 (REGION GA2) 



0-5 ENTORHINAL AREA LEADS 

 - 20 MSEC *• 



aA .. A/V>A., 



05 

 SEC 



-0-5- 



FiG. 5 

 Cross-correlation functions in early training, prepared with an iBM-709 electronic computer. 

 Here, records from hippocampal zones CA., and CA, consistently lead those from the entor- 

 hinal area by 20-35 nisec, thus resembling the effects of septal stimulation in acute experiments 

 (see text). 



irregularity in frequency ot the hippocampal slow waves, with some 

 strong periodic activity at 3-4 cycles/sec. In cross-correlations, phase 

 angles of this slower 'resting' activity showed considerable variation, but 

 were most consistently those in which the dorsal hippocampus led the 

 entorhinal area. Examination ot this 'resting' 3-4 cycles/sec. activity in 

 the trained animal thus showed a strong resemblance to the phase 

 relations of the faster 5-6 cycles/sec. bursts in the approach performance 

 of the untrained animal. It is of some interest that these phase angles can 

 be apparently sharply reversed in relation to the performance of a trained 



