very Small Portions of Time. 323 



in a case of the kind. The probable error of the mean value of 

 successful series amounted to only ^~ ff dth part of the whole 

 value. The difference between the measurements in which dif- 

 ferent points of the nertf e were excited was, on account of the 

 shortness of the nerve, also very small, from one to two thousandths 

 of a second ; it was, however, ten times as great as the probable 

 error of the results of the measurements. The most probable 

 value of the velocity of propagation in the motor nerves of the 

 frog I found to be 26*4 metres, about eighty feet per second. 

 This quantity is indeed unexpectedly small, more than ten times 

 less than the velocity of sound in the air. 



For warm-blooded animals the method described is not appli- 

 cable, because it requires series of measurements which occupy 

 from one to two hours, during which the state of the body expe- 

 rimented with must remain constant. I have therefore had an 

 apparatus with a rotating cylinder constructed by M. E. Rekoss, 

 with which I have made the first trial experiments on frogs, and 

 which may perhaps be made use of with warm-blooded animals. 

 The principle of the instrument is not quite the same as in the 

 apparatus of Siemens. The glass cylinder, constructed with 

 great exactness, stands vertical ; for the purposes of experiment 

 its surface is covered with a thin coating of lampblack ; against 

 this a point can be made to press ; the point is attached to a 

 lever which is connected with the muscle, and when the latter 

 contracts, the point is elevated. As long as the point remains 

 at the same elevation, it simply describes a horizontal circle 

 round the rotating cylinder. If the cylinder stand still and the 

 muscle contract, a vertical line is drawn upon the surface of the 

 cylinder ; but if the cylinder rotates during the contraction of 

 the muscle, a curve which first ascends and afterwards descends 

 is produced. If two different points of a nerve be excited suc- 

 cessively, and if the moments of excitation coincide exactly with 

 one and the same position of the point upon the rotating cylin- 

 der, two congruent curves are produced, which, however, appear 

 moved towards each other in a horizontal direction. The mag- 

 nitude of the displacement corresponds to the time of propaga- 

 tion in the length of nerve between the two points of excitation. 

 In this case, also, each single experiment shows whether the 

 duration and strength of the contraction were equal in both 

 instances. If this be the case, the two curves are congruent ; 

 if not, incongruent. Thus each single experiment here takes the 

 place of a whole series of experiments according to the former 

 process ; but it must be confessed, that, up to the present time, 

 I have not attained the same degree of exactness and agreement 

 in the results. 



How stands the question in the case of man ? We must expe- 

 riment on man under much more complicated conditions than 



