TIME-KEEPERS

TIME-KEEPERS, or Instruments for measuring Time. See CLOCK, DIAL, &c.

TIME-KEEPERS for finding the Longitude. Gemma Frisius seems to have been the first who suggested the method of finding the longitude at sea by means of watches or time-keepers; which machines were then, as he says, but lately invented. After him Metius, and some others, attempted it; but the state of watch-making was then too imperfect for this purpose. The

dispute between Hooke and Huygens, concerning the invention and application of the pendulum-spring to watches, was long and violent: each of them claiming this curious and most useful invention, and representing the other as a pirate. It is probable that their claims were each of them just; nor is this the only instance in which different persons have made the same discovery, nearly about the same time. Hooke and Huygens, each of them, on making this discovery, applied it to the purpose of discovering the longitude at sea. Some disputes, however, between the former of those gentlemen and the English ministry at that time, prevented the making any experiments with watches constructed by him; but many experiments were made with watches constructed by Mr Huygens, from which it appeared that those watches were of no real use at sea for this purpose. Dr Hooke never, as far as we know, made a full discovery of his inventions of this kind; but many hints are dropped in different parts of the Philosophical Transactions, his Philosophical Collections, and Cutlerian Lectures; of which later mechanics have undoubted availed themselves.

In 1714, an act passed for giving 20,000l. to that person who should first discover a method by which a ship might sail from England to any port in the West Indies, without having committed an error of 30' in her longitude, on arriving at the said port. The first who turned his thoughts this way, in consequence of this public encouragement, was Henry Sully, an Englishman, but who had left England before the passing of this act: for in 1714 he printed, at Vienna, a small tract on the subject of watch-making; and soon after he removed from thence and settled at Paris, where he spent the remainder of his life in improving time-keepers for the discovery of the longitude. In 1716 he presented a watch of his own making to the Royal Academy of Sciences, which was much approved. It is particularly said, that he had greatly diminished the friction; and that what he had not taken entirely away, he had by a very singular address rendered uniform. He went to Bordeaux in 1726, for the convenience of trying his watches, and died there in 1728. The greater part of what is yet known of watch-making in France is principally to be attributed to him; for the famous Julien le Roy was his pupil, and owed most of his inventions to him, which he afterwards perfected and executed; and this gentleman, his son, and M. Berthoud, are the only persons in France who have turned their thoughts this way since the time of Sully. Several watches made by the two latter artists have been tried at sea at the expence of the king of France, and very voluminous accounts of these trials have been published with great pomp; but the facts which are there related are so very few, and those few enveloped in such a volume of words, vague and indeterminate in their meaning, that it is scarcely possible to discover from thence what these watches are capable of performing.

M. Berthoud, in a pretty bulky pamphlet in 410, has with great labour collected together a few of the principal facts which resulted from the three last trials that were made of two time-keepers constructed by M. Le Roy, marked A and S; and of two of his own construction, denominated N^o 6. and N^o 8. which are as follow:

In the month of October 1771, two watches made by M. Le Roy, marked A and S, and M. Berthoud's, N^o 8. were again sent out on trial under Messrs. Verdun, Borda, and Pingre. A was the same watch which had been tried before by the Marquis de Courtauld and M. Cassini; but that marked S was a new one. They had also with them a small watch made by M. Le Roy, which, on account of its size and form, they called La petite ronde; but this did not answer at all. The performances of the other three were as follow:

On the 17th of March the ship struck on the Wil-

mington Rock, which lies off the island of Antigua; and the thermometer of compensation for heat and cold of the watch A was broken by the shock, and the watch put entirely out of order. This accident was the cause of their putting back to Port Royal.

From this account it would appear, that M. Berthoud's time-keepers greatly exceed those of M. Le Roy: but it ought perhaps to be observed, that this (N^o 8.) is the only one of his making which has performed so well; and even this, on the former trial, did not go with any very great degree of regularity.

About the year 1726, Mr John Harrison, whose name is now so well known on account of his time-keepers, began to apply himself to the construction of them: and in the year 1736, one of them was tried on board his Majesty's ships, in a voyage to and from Lisbon; in which trial it gave so much satisfaction, that he received public encouragement to proceed, and began to entertain hopes of obtaining the reward offered by the act of the 12th of Queen Anne; in order to which, he made three other time-keepers, every one of which was more accurate, and better adapted to the purpose of measuring time truly at sea, than the former. The second of these was finished in 1739; and during the next ten years, its going was so much admired by the ingenious men of those times, that the annual prize-medal, distributed by the Royal Society, for improvements in experimental philosophy, was given to Mr Harrison on St Andrew's day, 1749. Mr Harrison did not finish his third machine until the year 1758; having then a fourth in considerable forwardness, and which he finished in October 1761: this proved so much to his satisfaction, that he wrote immediately to the commissioners of the board of longitude, informing them that he was then ready to make the ultimate trial prescribed by the above-mentioned act. Accordingly Mr William Harrison, son of the inventor, embarked on board his Majesty's ship Deptford, in November 1761, with this fourth time-keeper, on a voyage for Jamaica; and the longitude of the island, as shown by the time-keeper on his arrival there, differed but one minute and a quarter of the equator from the true longitude deduced from astronomical observations. The time-keeper also pointed out the longitudes of the several places which they saw in the course of the voyage, in a very exact manner. Mr Harrison junior returned to England with the time-keeper, in the latter end of March 1762, and found that it had erred in the whole, from its setting out to its return to England, no more than $1^\{\backslash circ\}\; 54\text{'}\; \backslash frac\{1\}\{2\}$ in time, or $28\backslash frac\{1\}\{2\}$ minutes of longitude.

Mr Harrison now claimed the whole reward of 20,000^l. offered by the act of the 12th of Queen Anne (1714); but some doubts arising in the minds of the commissioners concerning the true situation of the island of Jamaica, the manner in which the time at that place had been found, as well as at Portsmouth; and it being further suggested by some, that although the time-keeper happened to be right at these two times, namely when at Jamaica and on its return to England, it was by no means a proof that it had been always so in the intermediate times, another trial was proposed in a voyage to the island of Barbadoes, in which precautions were taken to obviate as many of those objections as possible. Accordingly, the commissioners having

ving previously sent out proper persons to make astronomical observations at that island, which, when compared with other corresponding ones made in England, would determine beyond a doubt its true situation; Mr William Harrison again set out with his father's time-keeper, in the latter end of the month of March 1764, the watch having been compared with equal altitudes before he set out at Portsmouth; and arrived at Barbadoes about the middle of May: where, on comparing it again with equal altitudes of the sun, it was found to show the difference of longitude between Portsmouth and Barbadoes $3^\{\backslash text\{h\}\}\; 55\text{'}\; 3\text{'}\text{'}$. The true difference of longitude between these places, resulting from astronomical observations, is $3^\{\backslash text\{h\}\}\; 54\text{'}\; 20\text{'}\text{'}$: consequently the error of the watch was $43\text{'}\text{'}$, or $10\text{'}\; 45\text{'}\text{'}$ of longitude. The watch gained at the rate of $2\text{'}\text{'}$, 58 a-day on mean time, from February 29th to March 21st, at Portsmouth; and lost at the rate of $2\text{'}\text{'}$, 8 a-day, from May 14th to May 17th, at Barbadoes.

In consequence of this and the former trials, Mr Harrison received a moiety of the reward offered by the act of the 12th of Queen Anne, on his explaining the principles by which his time-keeper was constructed, and delivering it, as well as the former three, up to the commissioners of the longitude for the use of the public. He was also promised the other moiety of the reward, when other time-keepers were made on the same principles, either by himself or others, which performed equally well with that which he had last made. This last time-keeper was also sent down to the royal observatory at Greenwich, to be tried there under the direction of the Rev. Mr Maskelyne, his Majesty's astronomer royal. It did not appear, however, that during the time of this trial the watch went with the regularity that was expected; nor indeed with any thing near the regularity that it must actually have gone with during the course of the two voyages that had been made with it; which surprised many, and conveyed no favourable impression of the general utility of this method of discovering the longitude at sea; as it gave reasons for apprehending that the performance, even of the same watch, was not at all times equal; and consequently that little certainty could be expected in the performance of different ones. Moreover, the watch was now found to go faster than it did during its voyage to and from Barbadoes, by about 18 or 19 seconds in 24 hours: but this circumstance was accounted for by Mr Harrison, in a publication intitled, Remarks on a Pamphlet lately published by the Rev. Mr Maskelyne; where he tells us, that not expecting the watch would be required of him so soon as it was, he had altered the rate of its going, by trying some experiments which he had not time to finish before he was ordered to deliver the watch up to the board. It is possible that the watch might be disordered by these experiments, and that disorder be the cause of its subsequent irregularity.

Soon after this trial, the commissioners of longitude agreed with Mr Kendall, one of the watchmakers appointed by them to receive Mr Harrison's discoveries, to make another watch on the very same construction with this, in order to determine whether other watchmakers could make them from the account which Mr Harrison had given, as well as himself. The event proved the affirmative: for the watch produced by Mr

Kendall, in consequence of this agreement, went considerably better than that which had been made by Mr Harrison himself: and indeed better than any which have been made since on other principles, this only excepted which is the subject of the account before us.

This watch, made by Mr Kendall on Mr Harrison's construction, was sent out in the second voyage which Captain Cook made towards the South Pole, and round the world, in the years 1772, 1773, 1774, and 1775, to be tried under the care of Mr Wales, who was employed by the board of longitude for that purpose: and it appears from his account, that this watch was losing at the rate of $\backslash frac\{1\}\{2\}$ of a second a-day, from March the 24th to April 25th 1772, at the royal observatory at Greenwich. August 1st 1772, at the island of Madeira, latitude $32^\{\backslash circ\}\; \backslash frac\{1\}\{2\}\text{'}$ N. longitude $17^\{\backslash circ\}$ W. it lost at the rate of $1\text{'}\text{'}$, 77 a-day on mean time. At the Cape of Good Hope, latitude $33^\{\backslash circ\}\; \backslash frac\{1\}\{2\}\text{'}$ S. longitude $18^\{\backslash circ\}\; \backslash frac\{1\}\{2\}\text{'}$ E. it gained at the rate of $1\text{'}\text{'}$, 2 a-day on mean time, from November 2d to the 14th, 1772; and the greatest variation between the rates of its going on any two days was $5\text{'}\text{'}$, 4. At Dusky Bay, in New Zealand, latitude $45^\{\backslash circ\}\; \backslash frac\{1\}\{2\}\text{'}$ S. longitude $166^\{\backslash circ\}$ E. the watch gained at the rate of $6\text{'}\text{'}$, 7 a-day, from the 6th of April to the 25th, 1773, and its greatest variation was $3\text{'}\text{'}$, 6 from any one day to any other in that time. The watch gave the longitude of the island of Madeira $17^\{\backslash circ\}\; 6\text{'}\; \backslash frac\{1\}\{2\}\text{'}\text{'}$ W. which, for aught that is yet known to the contrary, is the exact longitude of that place. It made the longitude of the Cape of Good Hope $18^\{\backslash circ\}\; 12\text{'}\; \backslash frac\{1\}\{2\}\text{'}\text{'}$ E. which is about 11' short of the truth; and the longitude of Dusky Bay, in New Zealand, $163^\{\backslash circ\}\; 47\text{'}\; \backslash frac\{1\}\{2\}\text{'}\text{'}$ E. or too little by about $2^\{\backslash circ\}\; 15\text{'}\; \backslash frac\{1\}\{2\}\text{'}\text{'}$. But we ought, perhaps, with the person who had this watch then under his care, to observe, that in the compass of these 13 months the watch had passed through all climates, from the latitude of $51\; \backslash frac\{1\}\{2\}^\{\backslash circ\}$ N. to $67^\{\backslash circ\}$ S. and over a space nearly equal to the whole equatorial circumference of the earth. The only defect which appears to have been in this watch is, that its rate of going was continually accelerated; but in the three years and a half that it was under this trial, it never amounted to $14\text{'}\text{'}\; \backslash frac\{1\}\{2\}\text{'}\text{'}$ a-day; for on its return to Greenwich, in the month of August 1775, it gained only $13\text{'}\text{'}$ a-day; and its greatest rate during the voyage was at Fayal, one of the western islands, where it gained at the rate of $13\text{'}\text{'}$, 5 a-day on mean time.

In consequence of the going of this watch, the House of Commons were pleased, in 1774, to order the other moiety of the reward offered by the act of the 12th of Queen Anne, to be given to Mr Harrison: and to enact, that any other person who, by means of a time-keeper, the principles of which had not then been made public, should enable a ship to keep her longitude during a voyage of six months, within 60 geographical miles, or a degree of a great circle, should be entitled to a reward of 5000l.; that in case he could enable her to keep her longitude for the same time within 40 geographical miles, or two-thirds of a degree of a great circle, he should be entitled to a reward of 7500l.; or to a reward of 10,000l. if he enabled her to keep it for that time, within 30 geographical miles, or half a degree of a great circle. This determination, however, ought clearly to have been in minutes of the equator; as it now is, it may be subject to numberless disputes. This

Time. This reward was in 1780 claimed by Mr Arnold, who invented a new balance spring, with a compensation for the effects of heat and cold in the balance.

It appears from this report of the going of Mr Arnold's watch, that the mean rate which it went at during the month of February 1779, was losing 0", 31 a-day on mean solar time: during the month of March, its mean daily loss was 1", 37; during the month of April, 1", 38; during the month of May, 1", 34; the month of June, 1", 47; July 0", 31; August 0", 55. In the month of September it gained, on mean solar time, at the rate of 0", 44 a-day; in October, at the rate of 0", 38; at the rate of 0", 04 in the month of November; and it lost at the rates 0", 50, 0", 68, and 0", 60, respectively, in the months of December 1779, January and February 1780. From hence it appears, that the parts of this machine which are to counteract the effects of heat and cold, are most exactly adjusted; and perform their office with all the regularity that can ever be expected.

It further appears, that Mr Arnold has very happily adjusted his balance, to go alike in the different positions that the watch may be put into: for we find, that when the watch was in an horizontal position, with the face upwards, it gained at the rate of 1", 72 a-day, on mean solar time; with the face downwards, it gained 2", 83: in a vertical position, with the hour XII. upwards, it gained at the rate of 0", 35 a-day; with the hour VI. highest, at the rate of 3", 85 a-day; with the hour IX. highest, at the rate of 0", 29 a-day; and with the hour III. highest, it lost at the rate of 0", 35 per day.—The greatest difference between the rates at which the watch went on any two days in these 13 months, is 6", 69; namely, between its rates on October 8th and December 26th. The greatest difference between its rates of going on any day, and the next to it, is 4", 11; namely, between the 26th and 27th of December. So that the greatest error that it would have committed in the difference of longitude on any one day, would have been very little more than one minute; which, as Mr Arnold justly observes, is determining the longitude daily, to as great precision as the latitude can in general be determined.

If we take the mean rate which it went at during the month of February 1779, as a standard rate with which we may compare its going for the following 12 months, we shall find that the greatest error which it would have committed in the longitude shown by it, would have been 2', 33", 2, or 38' 18" in longitude: and this error happened about the end of six months, or in the beginning of September; for during these six months, the watch had all along gone slower than it did in the month of February, with which rate of going it is compared; but about the beginning of September it began rather to go faster than it did in the month of February, and by that means began to lessen its total error. And it continued to do so until the latter end of November, when it began again to go slower than it had done in the month of February, and of course to increase the quantity of its total error. And this it continued to do until the latter end of February 1780, when the error appeared again to be at a maximum, and equal to 2' 6", 6 in time, or 31' 39" of longitude. After this time it rather decreased to the end of the month.

Harrison's Time-keeper. See LONGITUDE, and the preceding article.