NOTES AND ADDITIONS. 313 



on a similar idea. Berfc made a wound in the back of a rat, 

 cut off a small piece of the end of the tail, and fixed the tail 

 firmly in the wound on the back. The tail of the rat coa- 

 lescing Avith the flesh of the back, it was attached at two points 

 like the handle of a pot. The original root of the tail was 

 then cut through, so that the attachment to the back alone 

 remained. If the free end of the tail, which was originally 

 the tail-root of the rat so treated, is pinched, the animal 

 feels it ; so that the irritation is evidently transmitted in 

 the sensory nerves in a direction opposite to that which is 

 usual in the tail of a rat under normal conditions, and it is 

 accordingly evident that the sensory nerves of the tail have 

 the power of transmitting the excitement in both directions. 



12. Negative Yariation and Excitement (p. 220). 



That negative variation is a constant and inseparable 

 accompaniment of nerve-excitement has been shown by 

 du Bois-Reymond by a large number of caieful and varied 

 experiments, which have been confirmed and extended in 

 various directions by many observers. It makes no difference 

 by what irritant the nerve is excited ; and both motor and. 

 sensory nerves are conditioned exactly alike in this matter. 

 From a large number of experiments to select but one of 

 peculiar interest, I may allude to the experiment recently 

 made with the nerve of sight. If the eye is extracted and 

 prepared in connection with a portion of the nerve of sight, 

 and if the latter is suitably tested as to its nerve-current, 

 and light is then allowed to fall on the eye, previously shaded, 

 then the current of this nerve exhibits negative variation. 



If ligatures are applied to a nerve so that the excitement 

 can no longer propagate itself from one side to the other, 

 irritation of one side causes no negative variation in the 

 other side. This experiment is of importance because it 

 affords a means of proving with sufficient certainty that no 



