THE CHRONOMETER

The British Government in 1713 offered a reward of £20,000 for any method by which the longitude could at all times be determined at sea; the whole reward would be given if the method, when tested by a voyage to the West Indies, were found to be within 30 miles of the truth, £15.000 if within 40 miles, and £10,000 if within 60 miles of the truth.

John Harrison (1693—1776), a Yorkshire carpenter, came to London in 1728 with drawings of a watch which he thought would earn the reward. In 1735 he presented his first watch to be tested by the Board of Longitude. He was sent to Lisbon with it, when he corrected the dead reckoning about I degrees.

The master of H.M.S. Oxford on the homeward voyage reported the landfall to be the Start, but Harrison, trusting to his watch, insisted it was the Lizard, and he was found to be right.

In 1739 he completed his second watch, which was not tried at sea. In 1749 he presented his third watch, which gained the gold medal of the Royal Society. It was less complicated and more accurate than the second, having an error of only three or four seconds a week. He then constructed his fourth watch, and applied for the full reward.

The tests were very severe. The chronometer was compared in the Observatory at Greenwich, sealed up and sent to Portsmouth, where Robertson the mathematician found its error by equal altitudes, which he reported to the Admiralty.

The watch was put on board H.M.S. Deptford, Capt. Diggs. It was secured by four separate locks, the keys of which were held by Governor Lyttleton of Jamaica, Captain Diggs, the first Lieutenant, and Harrison's son.

The Deptford sailed 18th November, 1761. During the voyage to Madeira the chronometer corrected the dead reckoning 1 degrees. The ship arrived at Madeira three days before H.M.S. Beaver, which had sailed ten days before the Deptford-caused by trusting to dead reckoning. From Madeira to Jamaica the watch corrected the longitude to the extent of three degrees, while some ships differed from the correct longitude five degrees.

When the ship arrived at Port Royal the watch was found to be in error only five seconds. On the return voyage to Portsmouth the error in longitude was less than 18 miles.

In 1764 the same watch was sent to Barbados. He gave with the chronometer a temperature scale which if taken into consideration would have shown an error in 156 days of only 15 seconds. But it was not until 1773, after the King had personally interposed on his behalf, that he received the full reward of £20,000. He also received a further sum from the East India. Company.

The two chief features are the "fusee," the escapement, and the balance. The fusee has the effect of rendering what would be a variable force uniform. The escapement regulates the beat.

The balance expands or contracts with heat or cold. By an arrangement

in two segments, each composed of two metals, steel on the inside and brass on the outside, the unequal contraction or expansion is nullified and the chronometer is not unduly affected by temperature.

Although every care and device are used in the construction of chronometers it is not possible to make a perfect instrument; but every care should be used in handling them. They should be wound carefully at about the same hour each day. The key should always be turned until it is felt to "butt."

They should be kept in well-padded boxes, or, as the latest method is, with springs fitted to fillet-pieces. They should not be subject to any unusual movement, as a chronometer is easily stopped by a rotary motion; and care should be taken that they do not suffer any greater changes of temperature than are unavoidable.

A good stop-watch is a most valuable adjunct to the chronometer. The watch should be compared both before and after use with the chronometer. The chronometer is merely a very perfect watch, in which the balancewheel is so constructed that changes of temperature have the least possible effect upon the time of its oscillation; such a balance is called a compensation balance. A chronometer may be well compensated for temperature and yet its rate may be gaining or losing on the time it is intended to keep; the compensation is good when changes of temperature do not affect the rate.

It is not necessary that a chronometer's rate should be zero (or even very small, except that a small rate is practically convenient); it is sufficient if the rate, whatever it is, remains constant. The indications of a chronometer at any instant require a correction for the whole accumulated error up to that instant. If the correction is known for any given time, together with the rate, the correction for any subsequent time is known.

Winding.-Most chronometers are now made to run either eight days or two days. The former are wound every seventh day, the latter daily, so that in case the winding should be forgotten for twenty-four hours the chronometers will still be found running. But it is of importance that they should be wound regularly at stated intervals; otherwise an unused part of the spring comes into action, and an irregularity in the rate may result.

Chronometers are wound with a given number of half-turns of the key. It is well to know this number, and to count in winding, in order to avoid a sudden jerk at the last turn: still, the chronometer should always be wound as far as it will go, that is, until it resists further winding. This resistance is produced not by the end of the chain, but by a catch provided to act at the proper time and thus protect the chain.

When a chronometer has stopped it does not again start immediately after being wound up. It is necessary to give the whole instrument a quick rotatory movement, by which the balance-wheel is set in motion. This must be done with care, however, and with little more force than is necessary to produce the result; afterwards the chronometer must be guarded from all sudden motions. The hands of a chronometer can be moved without injury to the instrument, so that it may be set proximately to the true time. It is, however, not advisable to do this often.

Transporting.-Chronometers transported on board ship should be placed as near the centre of motion as possible, and allowed to swing freely in their gimbals, so that they may preserve a horizontal position. They should also be kept as nearly as possible in a uniform temperature.

When transported by land, the chronometer should no longer be allowed to swing in its gimbals, but is to be fastened by a clamp provided for the purpose; for the sudden motions which it is then liable to receive would set it in violent oscillation in the gimbals, and produce more effect than if allowed to act directly.

It has been found that the rates of chronometers have been affected by masses of iron in their vicinity, indicating a magnetic polarity of their balances. Such polarity may exist in the balance when it first comes from the hands of the maker, or it may be acquired by the chronometer standing a long time in the same position with respect to the magnetic meridian. In order to avoid any error that might result from this polarity (whether known or unknown) it will be well to keep the chronometers always in the same position; and they should not be removed from the ship to be rated; but their rates should be found after they are placed in the position they are to occupy.

Pocket Chronometers should be kept at all times in the same position : consequently, if actually carried in the pocket during the day, they should be suspended vertically at night.

Comparison of Chronometers.-One (supposed to be the best) instrument is selected as the standard, and with this all others are compared after winding. A record should be made and retained of the comparisons, which will furnish a graphic history of the performance of each instrument. For convenience, the standard may be distinguished by the letter R (reference), and the others by the letters A, B, etc., as far as they extend in number; thus—

The second differences, in the last column, are the daily rates, if the standard does not change.

The practical benefit of a system of daily comparisons is that a guard may be kept on the steadiness of the rates of the instruments.

Comparison by Astronomical Observations.-When one or more chronometers have to be regulated by means of astronomical observations, these observations are made with but one of them, and the corrections of all the others are found by comparing them with this. On board ship the chronometers are never brought on deck; but the observations are made with a watch (often called a “hack watch "), or it may be a pocket chronometer, which is compared with the chronometer either before or after, or both before and after, the observations. The double comparison is necessary where extreme precision is required, in order to eliminate any difference of the rates of the watch and chronometer.

Comparison by Coincident Beats.-When two chronometers are compared

which keep the same kind of time, and both of which beat half seconds, it will mostly happen that the beats of the two instruments are not synchronous, but one will fall after the other by a certain fraction of a beat, which will be pretty nearly constant, and must be estimated by the ear. This estimate may be made within half a beat, or a quarter of a second, without difficulty, but it requires much practice to estimate the fraction closer, and with certainty.

Rate for Temperature.-Each chronometer should be accompanied with a record from a responsible chronometer maker, or, better still, from an Observatory, showing the daily rates for mean temperatures for each 10°, say from 40° to 100° Fahr.; then with a maximum and minimum thermometer in the chronometer-case, the actual temperature of the preceding day is recorded as soon as the case is opened for winding in the morning. Then, referring to the tabulated record of observed daily rates according to temperature, the rate for the preceding day is found by inspection, and, applying this according to its sign to the sum of the accumulated daily rates up to the previous day, there will be found the whole amount of the accumulated rate on the given day, to be applied as a correction to the primary error. Although the rates may differ with lapse of time, etc., it is more likely that the differences of rates for corresponding temperatures will remain the same, or nearly so.

These remarks have been taken from Chauvenet's "Manual of Spherical and Practical Astronomy"; see also Admiral Shadwell's excellent work, "Notes on the Management of Chronometers."

Application of the Original Error and Daily Rate of a Chronometer.-The chronometers carried on board British ships show Greenwich date approximately. When put on board, before the voyage commences, the optician who has had charge of the instruments furnishes the error by which each chronometer differs from Greenwich time on the noon of the day, and also the rate per day (of loss or gain on Greenwich time) that each instrument has been ascertained to keep during the period of supervision over them.

The given error-generally termed the original error-may indicate that the chronometer to which it is applicable is fast or slow on Greenwich mean time, and will be specified accordingly.

The given rate-called the daily rate-will be indicated as to whether the chronometer is gaining or losing on Greenwich mean time.

The application of the error and rate to the time shown by chronometer, to get the correct Greenwich mean time, is a purely arithmetical calculation, and may be shown here.

I.

TO FIND GREENWICH MEAN TIME BY CHRONOMETER

I. Given an Original Error and a Daily Rate

Write down the chronometer time astronomically, with the month and day prefixed; under it write the original error, which subtract if the chronometer had been found fast, but add if found slow.

2. Find the number of days elapsed since the date the chronometer was rated; multiply this number by the daily rate, and the product will be the accumulated rate to be subtracted from the time if the chronometer had been found to gain, but to be added if it had been found to lose.

Example. A chronometer indicating 5h. 40m. 42s. on November 11th had been found 19m. 58s. slow on Greenwich mean noon on July 2nd and was gaining 45s. daily; find the Greenwich mean time on Nov. 11th.

Or the error of the chronometer at the Greenwich mean time may be found, and applied directly to the chronometer time.

Find the accumulated rate as in (2) and place it under the original error; then if the error was fast and the rate was gaining, or if the error was slow and the rate was losing, the error of the chronometer would now be greater than before, hence add the two together and keep the same name. But if the error was fast and the rate was losing, or if the error was slow and the rate was gaining, the error of the chronometer would now be found by subtraction, taking care to change the name of the error, if the accumulated rate was the greater.

If, during the voyage, when calling at a port where there is a good observatory, or by astronomical observations at a place the position of which is well known, an error and rate are determined anew, the process of finding the Greenwich mean time is exactly on the basis of that just described; and according to the Rules here given.

II. Given Two Errors to find a Daily Rate

When two chronometric errors have been found, and, by means of the elapsed time, you have to find a daily rate, proceed as follows:

1. Write down the time by chronometer, with the month and day before it; under this time write the second error, adding it if slow, subtracting it if fast; the result will be the approximate Greenwich date.

2. For the Daily Rate.-Write one error under the other, then—

Both errors fast, or both slow, take their difference;
One error fast and the other slow, take their sum ;

the sum or difference must be converted into seconds, and divided by the number of days elapsed between the dates of the two errors. The result will be the daily rate, in seconds, or seconds and tenths, or perhaps tenths only.