322 Mr. R. Sabine on Measuring the Contour of 



exact results, the sensitiveness of such a means of indication 

 being more than a thousand times less than that of even a 

 Morse magnet. But this beautiful experiment was made 

 before Morse magnets were invented, and fifteen years before 

 a submarine cable was attempted, from which date exacter 

 knowledge has been acquired of the principles involved ; and 

 much more delicate methods have consequently been devised. 

 In all these methods, however, the same rule holds good : 

 some work has to be performed by the current, either in 

 moving a magnet or in neutralizing some given amount of 

 opposite electricity : and it is only when this performance is 

 completed that the indication is received. 



It is obviously impossible to divide the length of wire ope- 

 rated upon by the time required to indicate a signal, and to 

 call the quotient the "velocity of electricity" or even "the 

 velocity of a wave." Nor would any statement of time under 

 such circumstances be complete without all the data being 

 given of resistance and inductive capacity of the line, and of 

 resistance and of some relative constant of sensitivity of the 

 indicating instrument. In all the older experiments these data 

 were omitted, and we can only make an approximate estimate 

 of them in most cases. When we do so and reduce the results 

 to some imaginary line of unit resistance and unit inductive 

 capacity, we find the reduced times agree for those with similar 

 indicating systems. Where this was not the case, the reduced 

 times are found to correspond with the relative degrees of sen- 

 sitiveness of the indicating systems employed*. 



I have lately devised the following method, by which the 

 electrical state of any given point in a line may be examined 

 quantitatively at intervals of 10 I 00 of a second or less after 

 starting the electric impulse, so that the exact contour of a 

 wave, from the first instant of its gaining sufficient strength to 

 move a delicate mirror-galvanometer, may be measured from 

 point to point to its maximum, and, if required, to its gradual 

 disappearance. And by this method it becomes possible also 

 to ascertain the speed of a wave, by observing the time which 

 is taken for any given part of it to pass between two given 

 points in the line. 



This method is in principle as follows : — The conductor or 

 line, a b (fig. 1), is furnished at one end (a) with a contact- 

 key, e, which can be pressed upon the contact-stud g, and 

 puts in circuit the battery B, one pole of which is attached to 

 an earth-plate E ; at the other end (b) the conductor is put 

 to a second earth-plate (E'). Some intermediate point, c, is 



* A tabulated statement of the reduced times from all the older mea- 

 surements may be found in the ' Telegraph Journal ' (1873), vol. i. p. 186. 



