EXCURSIONS OF THE CAPILLARY ELECTROMETER. 
93 
hole is cut so that the photograph can be examined by transmitted light. The radius 
of the carrier is exactly equal to that of the pendulum on which the plate was 
exposed, and a fine wire or piece of horsehair is stretched from D to 0, passing close 
underneath the glass without touching it, which serves as a radial line of reference. 
An index fixed to the carrier B, passing over a scale upon A, enables small angular 
displacements of the carrier to be accurately measured, thus determining the time 
intervals, and the ordinates are found by laying a graduated rule upon the negative 
just over the radial line DO, using as a reference circle the edge of the photograph, 
which is always well defined. The length of the subnormal is found by means of the 
flat rod, E, which carries at one end the glass plate, F, shown on a larger scale in 
fig. 4. On this plate is ruled a fine line, a, cq, continuous with the edge of E, but 
broken for a distance of about 2 millims. at about the middle. Through the gap thus 
left passes the tangent line, b, exactly at right angles to a, cq. When this line b, at 
the part where it is intersected by a, cq, is placed as a tangent to the curve which is 
to be analyzed, at the point where it is cut by the radial line, DO, the length cut off 
by the rod, E, upon the graduated rule, C, which is permanently fixed at right angles 
to DO, is the subnormal to the curve at that point. The plate carrier is then shifted 
through an angle corresponding to a known interval of time, as determined from the 
time record upon the photograph, and the subnormal measured again for the part 
of the curve thus brought over the radial line, DO. This process is repeated at 
sufficiently close intervals throughout the curve, the corresponding ordinates being 
measured, taking the edge of the photograph as a reference circle. 
For the application of this method it is first necessary to analyze a normal curve 
produced under the exact conditions, as to resistance, of the experimental photo¬ 
graphs. With a suitable instrument, the subnormal is practically a constant multiple 
of y, the distance of the meniscus from its point of rest, and the value of this 
constant multiplier has to be found. 
The following example of the analysis of the normal curve of the electrometer used 
in the physiological experiments described by Professor Burdon Sanderson in a 
paper on the “ Photographic Determination of the Time-relations of the Changes which 
take place in Muscle during the period of so-called Latent Stimulation,” (‘ Proceedings,’ 
vol. 48, p. 14), will serve to show how this is done. In this particular case, as the 
curve was obtained for the purpose of analyzing the physiological photographs which 
we had taken, the circuit was led through the non-polarizable electrodes and the 
muscle exactly as it had been arranged for those experiments. An equivalent 
metallic resistance was then substituted for the preparation, and several other photo¬ 
graphs of the normal excursion were secured. But the resulting curves were found 
to give identical results and, therefore, the details of this one only are inserted here. 
First, the ordinates corresponding to time-intervals of •001 sec. were measured with 
an ivory rule, graduated very finely in millimetres, the tenth of a millimetre being 
