CANADIAN LONGITUDES. 453 



ON THE EMPLOYMENT OF THE ELECTRIC TELEGRAPH 

 IN DETERMINING THE LONGITUDE OF SOME OF THE 

 PRINCIPAL PLACES IN CANADA. 



BY LIEUT. E. D. ASHE, R.N., F.R.A.S. 



The introduction of the telegraph wire into an Observatory, and 

 the facihty and rapidity of registering observations by its means, may 

 be considered one of the most useful adaptations of the age ; and I 

 never recollect having been more deeply impressed by the idea of 

 man's intellectual development, than I was when I heard the " relay" 

 in the Observatory, Quebec, beating the seconds of the Sidereal clock 

 in the Observatory at Cambridge. I have long ceased to wonder at 

 the snorting locomotive as it dashes past at the rate of forty miles an 

 hour, dragging some hundreds of human beings in its wake. Man 

 can reason step by step, from the tea kettle until he arrives at the 

 steam engine, but when we hear the pulsations of a clock, be it ever 

 so far off, our reasoning faculties stagger under the stupendous fact. 

 It follows then, as a matter of course, that as the beating of a clock 

 in one observatory can be heard in another, no method can be so accu- 

 rate for determining the difference of meridians as the mode of doing 

 so by the electric telegraph. 



Two fixed observatories being connected by the telegraph wire, there 

 are various modes of determining their difference of meridians. Per- 

 haps the most accurate is to send a signal every time a star of a 

 pre-arranged list passes each wire of their respective transit instrument, 

 and as the time of these signals is carefully noted in each, it follows 

 that the time taken by each star to pass from the meridian of one 

 observatory to that of the other can be most accurately obtained. 

 Care, however, must be taken to change the observers in order to 

 eliminate the personal equation. And if the signal be sent from east 

 to west, and again from west to east, the time occupied by the signal 

 in passing along the line causes the meridional difference to be too 

 great in one instance and too little in the other, and consequently the 

 mean gives a correct result. For instance, let the observatory A be 

 twenty minutes to the eastward of an observatory B, and suppose the 

 signal to occupy one second in going along the line, then if A sends a 



