must be received with fome caution, as Mr. Barlow, in 1597, fays that in a perfonal conference with two Eaft Indians he was told by them, that inftead of our compafs they made ufe of a magnetical needle of fix inches or longer, fet upon a pin. in a difh of white China earth filled with water; that in the bottom of the difh they had two cross lines, to mark the four principal winds, and that the rest of the divifions were left to the fkill of the pilot. But to return to Europe, Mr. Perrault, in his parallel between the ancients and the moderns, has cited fome verfes of Guyot de Provins, who wrote in 1180, which fhew diftinctly that the mariner's compafs was known in the South of France at that time.

By moft writers the invention of the compafs is afcribed to Flavio Goin of Analfi in Campanee, who lived about the year 1300; and he is faid to be the first that applied it to navigation in the Medi

terranean.

Mr. de Lalande informs us, that in "Le tréfor de Brunet," a manuscript in the French king's library, there is a paffage which proves that the compafs was made ufe of about the year 1260.

Here however it may be obferved, that though a magnet, which has only two poles, will always, when freely fufpended, place itfelf in the magnetic meridian, or in the fame plane with other good magnets; yet when a magnet has more than two poles, thefe may be fo fituated that the magnet will not traverfe, that is, will have no directive power.

Thus fuppofe an oblong magnetic needle to have a north polarity equally ftrong at each end, and a fouth polarity in the middle; it is plain, that as each has an equal tendency towards the north, neither of them can be directed towards the north in preference to the other; confequently the nee

dle

dle cannot traverse. Though this cafe very feldom occurs, yet there are many others, where a needle, when fixed to a card on which the points of a compafs are drawn, may occafion confiderable errors; this has been clearly proved by Dr. Knight and Capt. Greaves. Mr. R. Walker, of Jamaica, has alfo clearly proved, that the only proper fhape for magnetic compafs needles, is that where the line of direction is in the edge of the bar; each end of the bar fhould be alfo pointed. But it will be needlefs for me to enter into the fubject, as Mr. Walker means to lay his own ideas before the public, in a work which will contain much other curious matter on this part of magnetifm.

OF THE VARIATION OF THE COMPASS.

Though the north pole of the magnet is, in every part of the world, directed nearly towards the north, yet it very feldom points exactly thereto, and confequently the fouth pole of the magnet feldom points towards the fouth. In other words, the magnetic meridian feldom coincides with the meridian of the place, but generally varies from it fome degrees eastward or weftward.

This variation is different in different places on land as well as at fea, and is alfo continually varying in the fame place. For instance, the varia tion is not the fame in London as at Paris, or as at the Cape of Good Hope; and the declination at London, or at any other place, is not the fame now as it was twenty years ago.

This variation is always reckoned from the north; that is, if the north end of a needle vary to the east of the north, the variation is faid to be eafterly; and if it vary to the weft, the variation is faid to be wefterly.

The uncertainty of the quantity of this va

riation

riation in different parts of the world, is a great impediment to the perfecting of navigation; and philofophers have earnestly endeavoured to investigate it's caufe, and, if poffible, to correct the errors it occafions. The research has hitherto been in vain, though I muft own, I think the fubject is in the hand of a perfon, who bids fair to difpel much of the darkness with which it has hitherto been furrounded; and from what I have feen of Mr. R. Walker's theory, it appears to me correct and judicious, and needs only to be understood to produce conviction. He has alfo contrived a compafs to afcertain the variation of the needle at fea without calculation.

Though the directive power of the compafs was applied to the purposes of navigation in the fourteenth and fifteenth century, it does not appear, that there were any apprehenfions during that time of it's pointing otherwife than due north and fouth.

The variation of the compass is faid to have been firft difcovered by Columbus, the latter end of the fifteenth century. But the first perfon who difcovered that it was real, and was the fame to all needles in the fame place, is generally allowed to be Sebaftion Cabot. This was about the year 1497.

After the variation was discovered by Cabot, it was thought, for a long time, to be invariably the fame, at the fame places, in all ages; but Mr. Gellibrand, about the year 1625, difcovered that it was different at different times, in the fame place.

From fucceffive obfervations made afterwards, it appears, that this deviation was not a conftant quantity, but that it gradually diminished, and at lait about 1657 it was found that the needle pointed due north at London, and has ever fince

been

been increafing to the weftward of the north. So that in any one place the variations have a kind of libratory motion, traverfing through the north to unknown limits eastward and weftward. The prefent variation at London is about 23 degrees.

Dr. Halley fuppofed that the earth has within it a large magnetic globe, not fixed within to the external parts, having four magnetic poles, two fixed, and two moveable, and by this he has endeavoured to account for the phenomena of the needle. His application of this theory to facts, is in many refpects inadequate, in all laboured and unnatural. Mr. Euler has fhewn, that he can. with two magnetic poles placed on the furface of the earth, account for all the phenomena as well as Dr. Halley with four; but his theory has also various imperfections.

The variation of the needle may be illuftrated by placing feveral touched needles round a magnetic bar, fee fig. 12, pl. 2. Now if the earth be a great magnet, or if it have only a magnetic atmofphere, it is clear from this experiment that magnetic needles placed on it's furface would have different directions in different places, which is conformable to experience; and the apparent irregularities in the variation of the needle must be occafioned by the fituation of the magnetic poles of the earth.

If the magnetic poles agreed with thofe of the earth, there would be no variation, and the magnetic needle would point to the true north and fouth. If the axis of the magnetic poles paffed through the center of the earth, it would be eafy to affign the quantity of the variation at every place; but as this is not the cafe, to account regularly for the variation, it is neceffary to know the exact fituation of the magnetic poles of the earth, their number, force, and diftance from the

real

real poles; whether they hift their place, and if they move, the quantity of motion every year.

OF THE DIURNAL VARIATION OF THE NEEDLE.

About the year 1722, and 1723, Mr. George Graham made a number of obfervations on the diurnal variations of the magnetic needle. In the year 1750, Mr. Wargentin took notice of the regular diurnal variation of the needle; and alfo of it's being disturbed at the time of an aurora borealis. About the latter end of the year 1756, Mr. Canton began to make obfervations on the variation, and in 1759 communicated feveral valuable experiments to the Royal Society.

The obfervations were made by him for 603 days; on 574 out of thefe, the diurnal variation was regular. The abfolute variation of the needle weftward was increafing, from about eight or nine o'clock in the morning, till about one or two in the afternoon, when the needle became stationary for some time; after that the variation westward was decreafing; and the needle came back again to it's former fituation in the night, or by the next morning.

The diurnal variation is irregular when the needle moves flowly eastward in the latter part of the morning, or weftward in the latter part of the afternoon; alfo when it moves much either way after night, or fuddenly both ways in a fhort time. These irregularities feldom happen more than once or twice in a month, and are always accompanied with an aurora borealis. The diurnal variation in the months of June and July, is almost double that of January and December.

Mr. Canton supposes that the diurnal beat of the fun acts upon the magnetic parts of the earth, or rather upon the magnet included in the earth.

But