578 BRAIN MECHANISMS AND LEARNING 



both dorsal and ventral parts of the arch, and in the adjacent entorhinal 

 area of the pyriform cortex. 



Two essentially opposing hypotheses have been offered in explanation 

 of the major neural fluxes through the hippocanipal arch (Fig. i). These 



MPLANTED ELECTRODE PLACEMENTS 



DENTATE 

 GRANULE CELLS 



HIPPOCAMPAL 

 PYRAMIDAL 

 CELLS : 



B. SUGGESTED PATHWAYS FROM PHASE MEASUREMENTS 



(II UNTRAINED ANIMAL 



(2) TRAINED ANIMAL 



Fig. I 



A. Typical impLmtcd electrode placements in the dorsal and ventral parts of the hippocampal 

 arch (zones CA.„ CA3, CA4 dorsally, and CA., and CA4 ventrally), and m the amygdala 

 (AMYG) and entorhmal area (ENT). 



B. In the untrained animal, phase measurements with an electronic computer of the slow- 

 wave trains during approach performance are consistent with activity passing from septum 

 to dentate granule cells (DENT. GR), thence to hippocampal pyramidal cells (HIPP. PYR), 

 and across die subiculum (SUB) to the entorhinal area (ENT). This accords with pathways 

 suggested by Elliot Smith and Herrick. The ventral hippocampus shows little correlation of 

 itsslow-wave activity with the approach after the first few trials. This arrangement is reversed 

 in the trained animal, with a phase sequence in the slow-wave trains from entorhinal area to 

 hippocampal pyramids, thus following the temporo-ammonic tracts of Cajal to the dendritic 

 trees of the hippocampal pyramids. F, fornix and fimbria. 



classical accounts have been based primarily on morphological considera- 

 tions. Cajal (1909) suggested that activity arising m the entorhinal area 

 traversed the hippocampus by trans-synaptic activation of the pyramidal 

 cell layer, with an efferent flux into the fornix. Contrary hypotheses 

 advanced by Elhot Smith (1910) and Herrick (i93 3) h-ive proposed an 



