RALPH S. LILLIE 139 



properly adjusted, the penetration is complete; and the wave, on 

 emerging into the completely transmissive region beyond, recovers 

 its former velocity and continues to the end of the wire. An activa- 

 tion wave will in this manner penetrate two regions of decrement 

 separated by an interval of complete transmissivity, while it will be 

 blocked by a single continuous region of a length equal to the sum of 

 the other two. Such an experiment is closely comparable with Adri- 

 an's well known experiment in nerve.^^ 



^ It is, however, difficult, on account of the variable behavior of 

 different iron wires, to reproduce this result at will with the above 

 method. The following procedure is more satisfactory and gives the 



^ A and C IL 5 JJ 



a^t^^;^!iifff^a;;^/WF/--y-^;jv.^MfcfMH 



y 



A rj^ B 



Fig. 2. Form and arrangement of wires in the experiment showing the differ- 

 ence in transmissivity between continuous and discontinuous decrement areas. 

 See text. 



result just described in a large proportion of cases. Instead of pro- 

 ducing as above a region of decrement in a single continuous wire, 

 three separate lengths of wire are used, bent in one plane as in the 

 diagram (Fig. 2). Good contact between any two wires can be se- 

 cured by resting one bent end across the other so that the right- 

 angled pieces are in contact and the long portions in a straight line. 

 In such an arrangement with two completely transmissive wires an 

 activation wave travels from the end of one wire to the end of the other 

 without appreciable retardation at the junction, and the two wires 

 behave like a single wire of double length. The experiment is per- 

 formed as follows. The three lengths of wire (first rubbed bright and 

 clean) are passivated in the usual manner and transferred to a dish 

 containing acid of 65 to 70 per cent strength ; they are now in a com- 



