The telegraph manual: a complete history and description of the semaphoric, electric and magnetic telegraphs of Europe, Asia, Africa, and America, ancient and modern.

640 ATLANTIC OCEAN TELEGRAPHY.
I am also informed that the currents through the cable,
even immediately after it was submerged, were so weak that
relays were useless, and that not one perfect message was recorded by them, everything that was received being read from
the reflections of a galvanometer.
By comparing the above data with those of the new cable
now making by Messrs. Glasse and Elliott, for the Electric
and International Telegraph Company, the amount of current
which entered the 1,000 miles of cable when disconnected at
one end should not have exceeded 2 or 2.5 parts, instead of 7.5
and 8.5 parts.
The inference by rough calculation, therefore, is that there
was a fault offering a resistance equal to 1,000 or 1,200 miles
of cable, situated at a distance about 560 miles from one end
of the 1,200 mile coil on board the Agamemnon.
This, however, cannot be the fault first alluded to, situate
at about 270 miles from Valentia, but may have been the one
which caused such alarm when the ships were 500 miles from
Ireland, and when the signals ceased altogether and never
certainly recovered.
It is not at all improbable that the powerful currents from
the large induction coils have impaired the insulation, and that
had more moderate power been used, the cable would still
have been capable of transmitting messages.
To sati.fy myself on this point, I attached to the cable a
piece of gutta-percha covered wire, having first made a slight
incision in the gutta-percha to let the water reach the wire;
the wire was then bent so as to close up the defect. The defective wire was then placed in a jug of sea water, and the
latter connected with the " earth." After a few signals had
been sent from the induction coils into the cable, and, consequently, into the test wire, the electricity burnt through the
incision, rapidly burning a hole nearly one tenth of an inch in
diameter.
When the full force of the coils was brought to bear on the
test wire by removing them from the cable, and allowing the
electricity only one channel-viz., that of the test wire, the
discharges, as might be expected, burnt a hole in the guttapercha under the water, half an inch in length, and the burnt
gutta-percha came floating up to the surface.
The foregoing experiments prove that when there are imperfections in the insulating covering, there is very great danger
arising from using such intense currents.
The size of the present conducting strand is too small to
have worked satisfactorily even had the insulation been sound.