MECHANICAL RESPONSE OF NERVE 515 



vibrations of a busy city. But when both the nerve- 

 chamber and the Optical Lever are fixed to the same 

 supporting-rod, relative movements, due to external disturb- 

 ance, are practically eliminated. This common supporting- 

 rod may be screwed securely to a wall. With these precautions, 

 I have been able to take records, without the least dis- 

 turbance from the adjacent electric tram line. As a matter of 

 fact, when the magnification required is only of a few hundred 

 times, nothing but gross carelessness could allow any source 

 of disturbance to remain. It is only when the magnification 

 has to be pushed to the order of a hundred thousand that 

 unusual care is necessary to avoid errors of disturbance. 

 One precaution which should, however, be taken, is that 

 arising from disturbance of the mirror by convection 

 currents of air. The remedy for this is obvious, namely* 

 a suitable glass cover. 



This is the order of magnification which is necessary for 

 the recording of response under a degree of stimulation 

 usual in making observations of excitatory electrical variation 

 with a very sensitive galvanometer. But while the sensitive- 

 ness of the galvanometric method of detecting response is 

 here nearing its limit, that of the mechanical method is in its 

 first stage only, and how greatly the sensitiveness of the 

 latter may be exalted when required will be shown in the 

 next chapter. 



I shall now explain how easy it is to study the physio- 

 logical variations induced in the animal nerve under various 

 agencies by means of the mechanical response. The following 

 experiments were performed on specimens of the sciatic 

 nerve of frog. A well-known reagent for abolition of ex- 

 citability of the nerve is ammonia. Its effect on mechanical 

 response is seen in fig. 316. In all the following experiments, 

 the stimulus applied was by fairly strong tetanising electrical 

 shocks, which were usually of two seconds' duration. Two 

 series of records were taken, successive responses being 

 recorded at intervals of one minute, before and after the 

 application of the chemical reagent. In fig. 316, the normal 



L L 2 



