Where failures have occurred in submarine telegraphs, the causes are now well understood and easily to be avoided.  Thus with the first Atlantic cable, its defects have all been carefully investigated by scientific men, and may be easily guarded against.  When this cable was in process of manufacture in the factory of Messrs. Glass, Elliot, & Co., in Greenwich, near London, it was coiled in four large vats, and there left exposed, day after day, to the heat of a summer sun, which was intensified by the tarred coating of the cable to one hundred and twenty degrees.  This went on, day after day, with the knowledge of the engineer and electrician of the company, although the directors had given explicit orders that sheds should be erected over the vats to prevent the possibility of such an occurrence.  As might have been foreseen, the gutta-percha was melted, so that the conductor which it was desired to insulate was so twisted by the coils that it was left quite bare in numberless places, thus weakening, and eventually, when the cable was submerged, destroying the insulation.  The injury was partially discovered before the cable was taken out of the factory at Greenwich, and a length of about thirty miles was cut out and condemned.  This, however, did not wholly remedy the difficulty, for the defective insulation became frequently and painfully apparent while the cable was being submerged.  Still further evidence of its imperfect condition was afforded when it came to be cut up for charms and trinkets.

The first cable was, to a great extent, an experiment,—­a leap in the dark.  Its material and construction were as good as the state of knowledge at that time provided, and in many respects not unsuitable; but the company could not avail itself, at that time, of the instruments or apparatus for testing its conducting power and insulation, in the manner since pointed out by experience.  The effects of temperature, as we have seen, were not provided for.  The vast differences in the conducting power of copper were discovered only by means of that cable, when made.  The mathematical law whereby the proportions of insulation to conduction are determined had not been fully investigated; and it was even argued by some of the pretended electricians in the employ of the company, that, the smaller the conductor, the more rapidly the current could pass through it.  No mode of protecting the external sheath from oxidation had then been discovered; and the kind of machinery necessary for submerging cables in deep water could only be theoretically assumed.

Looking back to that period, and granting that there was too much haste in the preparations, and that other mistakes were committed which could now be foreseen and avoided, it is not too much to say, that, if that cable could be laid and worked, as was done, after one failure in 1857, and the consequent uncoiling and storage of it in an exposed situation, and after three attempts in 1858, under the most fearful circumstances as to weather, it would be an easy task to lay a cable constructed and submerged by the light of present experience.