and a third throws the log-chip with the line attached over the ship's stern; the chip, floating upright, is kept stationary by the resistance of the water, while the vessel moves on, and the line runs out; the midshipman watches it until the limit of stray-line just passes the rail, when he sharply says "Turn"; the glass is quickly reversed, the sand begins to run into the lower compartment, and both time and space are reckoned from the word turn. The "knots" reel off slowly or rapidly according to the ship's velocity, and, when the last grain of sand runs out, the line is instantly stopped. The number of knots and tenths run out denotes the speed at the moment of making the experiment, and, according to the conditions of wind, sea, and sail for the whole hour, the speed is deduced for the hour, and so entered in the columns.
To draw in the line, a quick, strong jerk on it frees the plug, when the chip floats horizontally, and can be hauled aboard with little resistance.
The "course steered," which is always a coördinate entry with the velocity, is obtained from a standard compass, whose every error is found and tabulated, to be applied when necessary. As the course and velocity are entered every hour in the log-book, we have thus a continuous record of each direction in which the ship headed, together with the distance she proceeded in that direction.
The courses and distances are the data by which, with the aid of a traverse-table, the ship's position may be found at any time—the position by "dead reckoning," or "account," as it is called. Independently of this, the position—"by observation"—is daily found by the navigator by altitudes of the sun, the moon, or the stars.
Suppose a ship to leave New York at noon of any day, and that her "run" is accurately kept until noon of the next day, when the latitude and longitude by account are found. The ship may not really be in this position: currents may have borne her along or athwart her course, yet we can not discover them; they act on log-chip and vessel alike: but let the position "by observation" be determined for the same instant that it is "by account," and we have at once a standard of comparison whereby the treacherous streams are made known. If none exist, the position by the two methods should agree within the small limit of error due to the unavoidable imperfection of both observers and instruments.
A third mode of ascertaining the ship's run is by the patent log—an instrument constantly towed astern at the end of a long line. It has a small propeller which the motion through the water causes to revolve. This revolution is communicated to a series of cogged wheels connected with hands that point to a circular scale—an arrangement not unlike a gas-meter. Every noon the log is hauled aboard, read, reset, and then thrown overboard, to record again the number of miles by which the ship nears her port. Being entirely independent of both
