at the same time, by enforcing a stringent specification and constant supervision, to guard against any possible laxity in the details of its construction. The Cable so decided on weighed 35¾ cwt. per knot in air, but in water it did not exceed 14 cwt., being only a fraction heavier in that medium than the old Cable, though bearing more than twice the strain--the breaking strain of the new Cable being 7 tons 15 cwt. In water it was capable of bearing eleven miles of its own length perpendicularly suspended, and consequently had a margin of strength of more than four and a-half times that which was absolutely requisite for the deepest water. The core having been received from the gutta percha works, and carefully tested to note its electrical condition, was first taken to receive its padding of jute yarn, whereby the gutta percha would be protected against any pressure from the external iron sheath, which latter succeeded the jute. On former occasions this padding of jute had been saturated in a mixture of tar before being applied to the gutta percha; but experience had shown that this proceeding might lead to serious fallacies as to the electrical state of the core, cases having been repeatedly found where faults existed in the core itself--amounting to an almost total loss of insulation--which, however, were only discovered after being submerged and worked through, owing to the partial insulation conferred for a time upon the bad place by means of the tarred wrapping. The Atlantic core, therefore, was wrapped with jute which had been simply tanned in a solution of catechu, in order to preserve it from decay, and as fast as the wrapping proceeded the wrapped core was coiled into water, in which, not only at this stage, but ever afterwards until finally deposited in the sea, the Cable, complete or incomplete, was stored, and the water being able to freely pass through the tarred jute to the core, the least loss of insulation was at once apparent by the facility offered by the water to conduct away to earth the whole or a portion of the testing current.