10 National Geographic Magazine. 
the exchange of signals were connected directly to the ends of 
the line. 
Everything being ready, the routine of the work was as fol- 
lows :—The transit being carefully leveled was placed in the 
meridian by observation of zenith and circumpolar stars. From 
six to ten time stars, and two or three circumpolars were then 
observed, the instrument was reversed in the Ys and nearly the 
same number of stars observed in the new position. At some 
time agreed upon, generally when the. regular work of the tele- 
graph line was over for the day, the wires were connected up and 
one of the operators came to the observatory to assist in holding 
communication. By a simple arrangement of relays, in the line 
and chronograph circuits the chronometer at one station was 
made to register its second beats on the chronograph at the other, 
which was all the time being graduated into second spaces by its 
own chronometer. ‘This was done for about five minutes and the 
times of beginning and ending noted. Then the connections 
were reversed and both chronometers allowed to beat for five 
minutes on the chronograph at the first station. 
This method of exchanging signals was only practicable on 
land lines or very short cables. “The ordinary relay used on a 
land lime requires a strong current to work it, and would not be 
affected in the least by the delicate impulse sent over a long cable, 
consequently when the expedition came to compare chronometers 
over the 600 miles of cable between Aspinwall and Kingston, it 
was necessary to use another method. At that time the instru- 
ment in general use on submarine cable lines was what is known 
as Thompson’s mirror galvanometer. It consists of a coil of very 
fine insulated wire wound with great care on a spool or bobbin of 
vulcanite, about three inches in diameter and 14 inches thick. 
In a hole in the centre of the spool is made to slide a small tube, 
so that the end of the tube will be in the centre of the coil. In. 
the end of the tube is mounted a small mirror, swung in a vertical 
position on a single upright fibre of silk. Horizontally across the 
back of this mirror is secured a small permanent magnet, in length 
about the diameter of the mirror or about one-eighth to one- 
quarter of an inch. The mirror and magnet together weigh only 
one or two grains. When an electric current is sent through this 
coil it deflects the magnet and with it the mirror to the right or 
left. The apparatus is exceedingly sensitive so that it is influ- 
enced by very feeble currents. Communication has been main- 
