The next day there was a fresh cause of alarm. It was found that the Niagara had run some miles out of her course. Comparing the distance run by observation and by patent log, there was a difference of sixteen miles and a third. With such a percentage of loss, the cable would not hold out to reach Newfoundland. This was alarming, but it was soon explained. The mass of iron in the ship had affected the compass, so that it no longer pointed to the right quarter of the heavens. Had the Niagara been alone on the ocean, this might have caused serious trouble. But now appeared the great advantage of an attendant ship. It was at once arranged that the Gorgon should go ahead and lead the way. As she had no cable on board, her compasses were subject to no deviation. Accordingly she took her position in the advance, keeping the line along the great circle arc, which was the prescribed route. From that moment there was no variation, or but a very slight one. The two methods of computing the distance--by log and by observation--nearly coincided, and the ship varied scarcely a mile from her course till she entered Trinity Bay. It is not necessary to follow the whole voyage, for the record is the same from day to day. It is the same sleepless watching of the cable as it runs out day and night, and the same anxious estimate of the distance that still separates them from land. Communication is kept up constantly between the ships. Mr. Field's journal contains entries like these: "Saturday, July thirty-first. By eleven o'clock had paid out from the Niagara three hundred miles of cable; at forty-five minutes past two received signals from the Agamemnon that they had paid out from her three hundred miles of cable; at thirty-seven minutes past five finished coil on the berth-deck, and commenced paying out from the lower deck." "Monday, August second. The Niagara getting light, and rolling very much; it was not considered safe to carry sail to steady ship, for in case of accident it might be necessary to stop the vessel as soon as possible. Passed and signalled the Cunard steamer from Boston to Liverpool." Same day about noon, "imperfect insulation of cable detected in sending and receiving signals from the Agamemnon, which continued until forty minutes past five, when all was right again. The fault was found to be in the ward-room, about sixty miles from the lower end, which was immediately cut out, and taken out of the circuit."