From this period, having reached 2,000 fathoms of water, it was necessary to increase the strain by a ton, by which the rate of the cable was maintained in due proportion to that of the ship. At six o'clock in the evening some difficulty arose through the cable getting out of the sheaves of the paying-out machine, owing to the pitch and tar hardening in the groove,[20] and a splice of large dimensions passing over them. This was rectified by fixing additional guards and softening the tar with oil. It was necessary to bring up the ship, holding the cable by stoppers until it was again properly disposed around the pulleys. Some importance is due to this event, as showing that it is possible to "lay to" in deep water without continuing to pay out the cable, a point upon which doubts have frequently been expressed. Shortly after this the speed of the cable gained considerably on that of the ship, and up to nine o'clock, while the rate of the latter was about three knots, by the log, the cable was running out from five and a half to five and three-quarter knots. The strain was then raised to 25 cwt., but the wind and the sea increasing, and a current at the same time carrying the cable at an angle from the direct line of the ship's course, it was found insufficient to check the cable, which was at midnight making two and a half knots above the speed of the ship, and sometimes imperiling the safe uncoiling in the hold. The retarding force was therefore increased at two o'clock to an amount equivalent to 30 cwt., and then again--in consequence of the speed continuing to be more than it would be prudent to permit--to 35 cwt. By this the rate of the cable was brought to a little short of five knots, at which it continued steadily until 3.45 A.M., when it parted, the length paid out at the time being 380 miles. I had up to this attended personally to the regulation of the brakes, but finding that all was going on well, and it being necessary that I should be temporarily away from the machine--to ascertain the rate of the ship, to see how the cable was coming out of the hold, and also to visit the electrician's room--the machine was for the moment left in charge of a mechanic who had been engaged from the first in its construction and fitting, and was acquainted with its operation. In proceeding toward the fore part of the ship I heard the machine stop. I immediately called out to relieve the brakes, but when I reached the spot the cable was broken. On examining the machine--which was otherwise in perfect order--I found that the brakes had _not_ been released, and to this, or to the hand-wheel of the brake being turned the wrong way, may be attributed the stoppage and consequent fracture of the cable.