THE CIRCULATION OP CEREBRO-SPINAL SUBSTANCE. 577 



really intended to receive discharges of nervous energy from 

 the bulbous tips of axons? They outnumber the axonal end- 

 organs out of all proportion. Indeed, their multiplicity around 

 all the stems excepting the axon hardly points to them as ter- 

 minals endowed with such important functions as those at- 

 tributed to them. Again, Berkley states that "the twigs of 

 the dendrites and the fibers touch each other frequently and 

 in a manner that appears to be perfectly indifferent for the 

 different kinds of nervous substance, receptive and protective." 

 Such promiscuousness plainly testifies, it seems to us, against 

 the identity of the substances in contact being exposed sur- 

 faces capable of transmitting to each other a stream of nervous 

 impulses. 



In the light of our views, however, a function perfectly in 

 keeping with the experiments of Goddard in puppies, of Demoor 

 with morphine, of Berkley with alcohol, ricin, serum, etc., sug- 

 gests itself. We have seen that during functional activity the 

 gemmules project, while during inactivity they recede. If we 

 now connect these facts with the presence in the gemmules of 

 a thin layer of myelin immediately under their external or 

 limiting covering, and concede that the latter and the myelin 

 take part in the formation of each gemmule, it will become 

 evident that during the erethic state the surface of myelin ex- 

 posed to the action of the oxidizing substance of the plasma 

 will be greatly increased and the proportion of nervous energy 

 produced correspondingly augmented. Eetraction of the gem- 

 mules, on the other hand, by emptying them of their plasma, 

 will normally cause diminution of energy-production, the mye- 

 lin of the main channel sufficing to sustain nutritional functions 

 during sleep, for instance, when the gemmules are retracted. 

 We have what seems to us a counterpart of these minute struct- 

 ures in the muscle-cell, the myosinogen of the latter being 

 replaced by the myelin. We have also, in the processes out- 

 lined, an active and a passive functional stage in keeping with 

 other organs. All this so thoroughly coincides with the various 

 attributes which the gemmules have been shown by various in- 

 vestigators to possess, that we feel warranted in concluding that 

 the gemmules are peripheral extensions of the dendritic walls hav- 

 ing for their purpose to increase, when erect, the area of myelin 



