BAR 



BAR 



tube, as the mofl: efFeftual method of purging it of its air 

 and moitlure. The procefs is briefly this : he chufes a tube 

 of 2| lines or 3 hnes bore, and not exceeding half a line in 

 thicknefs ; he tills it with mercury within two inches of the 

 top, and holds it with the fealed end lowed in an inclined 

 poiition over a chafing dilh of burning charcoal, prefenting 

 iirlt the fealed end to the fire, and moving it obliquely over 

 the chafing-di(h. As the mercury is heated, the air bub- 

 bles appear hke fo many ftuds on the inner furface of the 

 tube, and gradually running into one another, afccnd to- 

 wards the higher parts of the tube, which are not heated ; 

 here they are condenfed and almoft difappear ; and after fuc- 

 cefTive emigrations, they acquire a bulk by their union, 

 which enables them at length to efcape. When the mercury 

 boils, its parts (Irike againft each other, and againll the 

 fides of the tube, with luch violence, that a perfon unac- 

 cuftomed to this operation is ready to apprehend their force 

 to be fufficient to break the tube. The mercui-y is thus 

 freed from all the heterogeneous particles contained in it, 

 together with their fuiTOiinding atmofpheres, and the air 

 which lines the infide of the tube, which cannot be eafily 

 expelled in any other way, is difcharged ; when this lall- 

 nicntioned fti-atum of air is thus expelled, the tube may be 

 afterwards emptied, and filled even with cold mercui-j-, and 

 will be found nearly 33 free from air as before. The mer- 

 cur)' in the tubes thus prepared by a determinate quantity 

 of heat, will rife higher than in thofe of the common fort, 

 and the barometers will more nearly correfpond with each 

 other ; whereas there will be a difference of lix or eight 

 lines in the afcent of the mercury in common barometers. 

 When this operation is completed, the mercury generally 

 remains fufpendcd at the top, and will not defcend to its 

 proper level without (liaking the tube to bring it down. 

 The tubes, which fliould be chofen not lefs than three feet 

 long, may now be filled to their proper length. 



Barometers of this kind rofe uniformly in a heated room ; 

 whilft the mercury in thofe that had been prepared in the 

 common way defcended, and in different proportions. When 

 the room cooled, the former defcended uniformly, and cor- 

 refponded with each other ; the latter rofe with the fame 

 irregularity with which they had before defcended, nor were 

 they found, at the clofe of the experiment, to ftand at the 

 fame relative heights as they did at the beginning of it. 

 The reafon of which is obvious, from the effects of heat on 

 the air remaining in unequal quantities in the tubes in the 

 one cafe, and on the purer mercury in the other. 



Another circumftance that requires attention in the con- 

 ftruction and ufe of baromieters is the temperature of the air; 

 for unlefs this remains the fame, the diinenfions of a given 

 quantity of mercuiy will be variable ; and the altitude of the 

 mercury wiU be an uncertain meafure of the weight of the at- 

 molphere, becaufe it is dilated by heat, and contratlcd by 

 cold, when probably its weight and prelTure are unchanged. 

 M. De Luc attended particularly to this circumllance, and 

 contrived to eilimate the eiietls of heat on the quickfilver in 

 the barometer, when it is ufed for accurate oblervations, by 

 means of a thermometer ; the fcale of which is divided in 

 fuch a manner as to indicate, with little labour of calcula- 

 tion, the correftion to be made on account of heat. As 

 an increafe of heat that is fufficient to raife the mercury in 

 the thermometer from the point of melting ice to that of 

 boiling water, will lengthen the cchimu of mercury in the 

 barometer fix lines, he divides each line in the fcale of th.e 

 barometer into four parts, each of which may be eafily fub- 

 dividcd into four lelTer parts, or fixtcenths of a line. The 

 fcale of the thermometer marking the interval between the 

 freezing and boihng points, and anfwering to the fix lines 



of the barometer, is divided into ninety-fix equal parts ; 

 each of which will conefpond to the lixteenth of a line in 

 the motion of the mercuiy in the barometer dilated by heat, 

 which mull be added to or fubtracled from the height of the 

 mercury in the barometer, for cveiy degree of the variatiori 

 of the thermometer fo graduated. A fcale of this kind, 

 continued above boiling or below freezing water, is annexed 

 to \vLi FofUibk Barometer, and Thermometer. M. de 

 Luc prepared two barometers with their refpedtive thermo- 

 meters graduated in the manner above explained ; he placed 

 one pair in the cellar of one houie, and the other pair in 

 tiie upper room of another houfe in a lower fituation, fo 

 as to be exaAly on a level with the cellar : he found that 

 the thermometer in the rjom rofe nine degrees, and the ba- 

 rometer -f-g of a line higher than thofe in the cellar ; whence 

 he fliews, that without allowing for the effett of heat, the 

 difference in the heights of thefe two barometers would have 

 indicated a difference of about forty-five feet in the heights 

 of thefe two places, though they were exaflly on the fame 

 level. M. De Luc's Rccherches, &c. vol. i. p. 193 — 199. 

 See Atmosphere, and the lequel of this article. 



M. Prins, an artifl in Holland, has made an improvement 

 in the refervoir of the fimple barometer, by means of which 

 the mercury contained in it is conllantly kept at the fame 

 level ; but the conlhuftion is difficult, and therefore it has 

 not been generally adopted. De Luc's Recherches, &c. 

 voh i. p. 3J. 



The common barometer is a kind of chamber barometer, 

 and ferves for obfeiving in a fixed place the changes of the 

 atmofphere ; but is not adapted for removal from one place 

 to another, and in this refpedl differs from the portable ba- 

 rometer. It is fometimes combined with a thermometer, 

 and fometimes alfo with an hygrometer, and in this form pre- 

 pared by the mathematical inftrument makers. An inllrument 

 of this kind conflrucled byMeffrs.Jones, opticians in London, 

 is exhibited \njig. 84, and confills of a barometer J, ther- 

 mometer aa, and hygrometer c, all in one mahogany frame. 

 The thermometer or hygrometer of this apparatus may be 

 conveniently feparated from the frame, and occaflonally ufed 

 apart, if it be neceffar^^. The thermometer is feparated by 

 means of two fcrews aa ; and the hygrometer, by unfcrew- 

 ing abrafs pin at the back of the frame. The index of the 

 hygrometer is fet at any time, merely by moving with the 

 finger the brafs wheel feen at c ; and the two fliding indexe* 

 of the barometer and thermometer are moved by rack-work, 

 fet in action by the key g placed in the holes h and /'. The 

 divifions cf the barometer plate b are in tenths of an inch, 

 from 28 to 31 inches; and thefe are fubdivided into hun- 

 dredths, by the Nonius or Vernier fcale, placed on a Hiding 

 flip of brafs, fimilar to that of the common barometers. 

 Tfiis Vernier (^. 85.) is divided into ten equal parts, all 

 of which are equal to eleven of thofe on the fcale of inclics, 

 or to eleven tenths. By this artifice, the height of the mer- 

 cury at E is evident merely by infpe£liou to the one hun- 

 dredth part of an inch. For underllanding this, it ftiould 

 be confidered that -Yiith part of-j^thof an inch is the-p^th 

 part of an inch. But every tenth of an inch in the fcale 

 B is divided into ten equal parts by the flip or Vernier A j 

 for fince ten divifions on that exceed ten on the fcale by one 

 divifion, that is, by one tenth of an inch, one divifion oa 

 the Vernier will exceed one divifion on the fcaie by one 

 tenth part, and two divillons on the Vernier Avill exCvcd two 

 on the feak- by two tenths, and lo on ; ihertfore every divi- 

 fion on the Vernier will exceed the fame number of di\ifion» 

 on the fcale by fo many tenths of a tenth, or by fo many 

 hundredth parts of an inch. Confequently the ten equal 

 divifions cf an inch on the fcale B lauil be confidered as fo 

 4 P 2 man/- 



