576 REroRT — 1882. 



liglit levers which are made to -write upon the same moving surface ; the muscle is 

 then stimulated at one end ; the wave of contraction, in passing along the fibres, 

 raises the levers in succession, and the space intervening between the commencement 

 of the curve described by the one lever and that of the other, gives the time taken 

 by the contraction-wave to traverse the portion of muscle between the levers. From 

 these data the velocity of the wave is readily calculated. 



It is found, however, that the measurements obtained by this method are of 

 much less value than those deducible from measurements of the rate of conduction of 

 an excitation along the muscular fibres, or of the rate of passage of the electric 

 changes which result from stimulation ; indeed, it is scarcely possible to obtain any 

 exactitude by its means, on account of the difficulty in determining exactly the 

 points at which the muscle-curves described by the levers begin. 



In order to obviate this difficulty I have caused the levers, instead of writing 

 directly upon a moving surface, successively to break galvanic circuits connected 

 with a Ruhmkorff induction-apparatus, the ends of the secondary coil being so 

 arranged that the sparks are transmitted through a sheet of smoked paper, moved 

 rapidly by means of a strong spring-myograph ; a time-tracing is at the same time 

 recorded on the paper by a chronograph. The interval between the pricks in the 

 smoked paper caused by the sparks, gives with great exactness the time occupied by 

 the contraction-wave in traversing the distance between the levers. 



The (vulcanite ) levers employed for the purpose are about six inches long and of 

 the third kind. They are so hinged at the fulcrum as to permit no lateral play, and 

 at a distance of about two inches from the fulcrum a thin piece of steel is rigidly 

 fixed so as to depend vertically ; the lower end of this rests on the muscle. 



Each lever carries at its free end a small semicircular bridge of copper wire, the 

 ends of which are bent downwards ; one end terminates in a platinum point, and the 

 other end is amalgamated. Immediately beneath the platinum point is a plate of 

 the same metal, capable of being raised up by means of a screw, so as to be brought 

 in contact with the point ; and beneath the amalgamated point is a small cup 

 containing mercury, the surface of which can also be raised by a screw (as in the 

 Helmholtz ' double-contact ' arrangement emploj-ed in the ' Frosch-unterbrecher ' 

 of du Bois-Reymond), so as to touch the amalgamated point, and by subsequently 

 depressing the mercurial surface, remain connected with it by a thread of mercury 

 only. J5y this contrivance any current which may be passing through tlie wire 

 from the platinum plate to the mercurial cup, or vice versd, is broken at the point 

 of contact with the plate tlie instant the lever is raised, but in consequence of the 

 breaking of the mercurial thread, the circuit is not closed again when the lever falls, 

 until the mercurial surface is again raised by the screw. 



In experimenting with this apparatus especial attention must be directed to the 

 regulation of the contact between the platinum point and plate, the latter being 

 screwed up with the greatest possible care, so as ouly just to make contact with 

 the point. 



The results which I have obtained by this method, although not yet sufficiently 

 numerous for detailed publication, have given a considerably higher velocity (between 

 2 and 3 m. per sec.) than those yielded by the original metliod of Aeb}'. 



A modification of the above method consists in the adoption of Pouillet's 

 method of time-measurement. The arrangements for this purpose are simpler, the 

 spring-myograph and the RuhmkorfT being dis]iensed with and their place taken 

 by a galvanometer. The second copper bridge, i.e. the one attached to the further 

 lever, is introduced directly into the galvanometer-battery circuit (by connecting 

 the wires of this circuit with its platinum plate and mercurial cup), but the first 

 copper bridge is so connected laterally as to short-circuit the current. The raising 

 of the first bridge, therefore, sends the current through the galvanometer, and the 

 raising of the second bridge breaks the circuit. During the interval between the 

 raising of the levers the current has been acting upon the galvanometer, and tht> 

 length of time it has been passing can be ascertained by noting the extent of 

 deflection of the galvanometer-needle. 



