244 INITIATIVE IN EVOLUTION 



system." The presence of certain special conditions, he points out, 

 causes the indifferent stimulus, which would otherwise be dispersed 

 in the higher centres, to be directed to a particular focus, and 

 eventually to lay down for itself a path to that part. A very interesting 

 detail of such a building of a new reflex is that " the stimuli from 

 which the new reflex is to be worked out shall be rigidly isolated." 

 Therefore to avoid any interference with the certainty of the experi- 

 ment, such matters as a personal bodily odour or kind of movement, 

 or even such a slight fact as a change in the mode of breathing 

 familiar to the dog on the part of the experimenter, has in the latest 

 experiments been removed by the application of the stimuli by 

 mechanical devices worked from another room, with results similar 

 to the earlier ones. Conditional reflexes can also be obtained from 

 stimuli arising from the locomotor apparatus, as the joints, eliminat- 

 ing the stimuli arising from the skin. Also certain parts of the 

 frontal lobes were extirpated and " when one part is extirpated the 

 reflex is obtained from the flexion of the joint, but not from the 

 skin ; if a different part be removed we can get the skin-reflex, but 

 not the reflex from the joint." He extirpated in one case the 

 greater portion of the posterior part of the brain and the dog lived 

 for several years after this in complete health. It was found easy 

 to obtain a conditional reflex for various intensities of illumination, 

 also for sound, and even a fine differentiation of tones. In another 

 dog the anterior half of the brain was removed and all the reflexes 

 before worked out in this animal disappeared, and yet in this 

 helpless condition of the dog he could train it to give that response 

 of the salivary glands which he called the " water-reflex," in which 

 fiist of all an irritating acid was introduced into the mouth and the 

 subsequent administration of water provoked an abundant secretion 

 of saliva which does not occur when water is poured into the mouth 

 of a normal dog. This was confirmed in another example in which 

 alone the centie for smell had been spared, and yet it was possible 

 in it to train the smell-reflexes also. I add one striking sentence 

 from Pawlow's address which, though an opinion, must be received 

 with the respect it deserves from such a source. " It is perhaps 

 not rash to think that some of the newly -formed conditional reflexes 

 can be transmitted hereditarily and become unconditional thereby." 



Indirect Evidence. 



From these limited but cogent pieces of evidence I turn to the 

 larger but confirmatory lines of indirect evidence and inference, 

 of which such works as those of Professors Sherrington, Bayliss, and 

 Starling, the notable address of Professor Macdonald at Portsmouth 



