BAR 



BAR 



The part HBCDI is filled with mercury, having a vacuum 

 above in A13. IF is filled with a light flwid, and FG with 

 another light flind, which will not mix with that in IF. 

 The cifteni G is of the fame diameter with AB. It is plain 

 that in this inflrument the ranpre of the feparating furface 

 at F mull be as much greater than that of the fiirface I, 

 as the area of I is greater than that of F ; and this ratio 

 may be felefted at plealure. This barometer is the beft of 

 thofe with an enlarged fcale ; it is mod delicately move- 

 able, and is the beft adapted to a chamber for the purpofe of 

 amiifement, by obfervations on the changes of the atmo- 

 fphcric preflure. It rifes or falls by tlie llighteft breeze, and 

 is continually in motion. The moil accurate method for 

 graduating fuch a barometer would be to make a mixture 

 of vitriolic acid and water, which fhould have -j-'^- of the 

 denfity of mercui-y. Then, let a long tube ftand vertical 

 in this fluid, and connect its upper end with the open end of 

 the barometer by a pipe with a branch to which the mouth 

 may be applied. By fucking through this pipe, the fluid 

 will rife both in the barometer and the other tube ; and the 

 rife of ten inches in this tube will correfpond to a defcent of 

 one inch in the common barometer. Thus every point of 

 the fcale may be adjufted in due proportion to the reft. 

 But nothing except adlual companion can determine what 

 particular point of the fcale correfponds to fome determined 

 inch of the common barometer. When this is done, the 

 whole becomes equally accurate. It is liable, however, to 

 fevcral inconveniences. Although the heights of the con- 

 tained fluids are always the fame in a conftant temperature, 

 neverthelefs their weight or prefl"ure on the bafe is not al- 

 ways the fame on account of the difference of their fpecific 

 gravity ; and though there be no fenfible difference in the 

 adion of thefe fluids againft the fides of the tube, yet there 

 is a continual action, and therefore the movements of this 

 barometer cannot be fo free as thofe of the fimple barome- 

 ter. Thefe differently coloured liquors mingle with one 

 another, and form a depofit on the fides of the tube, fo 

 that their rcfpective boundaries cannot always be afcertained 

 with precifion. The fluid of this barometer is alfo fubjetl 

 to evaporation ; and heat afts upon the fluids which it con- 

 tains. On account of thefe and fuch defefts, others have 

 had recourfe to an 



Horizontal or reSlangiilar barometer A BCD (fig. 8o. ) ; the 

 tube whereof is bent in form of a fquare BCD : at the top 

 of its perpendicular leg it is joined to a veffel or ciftern AB ; 

 and its variations accounted on the horizontal leg C D. 

 Now here the interval, or fpace of variation, may be made 

 of any extent at pleafure, and fo the minuteft; change 

 in the air become fenfible. For the diameter of the tube 

 CD being given, it is eafy to find the diameter of the vef- 

 fel AB, fo as that the fcale of defcent in the tube DC ftiall 

 have any given proportion to the fcale of afcent in the veiTel 

 AB ; the rule being that the diameter of the veffel is to 

 that of the tube in a fubduplicate reciprocal ratio of their 

 fcales. 



The diameters then of CD and AB being given, toge- 

 ther with the fcale of afcent of the mercuiy in the vefiel, 

 the fcale of mercury in the tube is found thus : as the fquare 

 of the diameter of the tube is to the fquare of the diam.eter 

 of the veffel, fo reciprocally, is the fcale of mercury in the 

 veffel, to the fcale of mercury in the tube. 



Caflini was the firft inventor of this kind of barometer, 

 though the fame conllruAion had been thought of, and 

 firft publiihed by M.J. Bernouilli, in the year 1710. 



This and the preceding contrivance of Huygens are 

 founded on a theorem in hydroftatics ; viz. that fluids, 

 having the fame bafe, gravitate according to their perpen- 



dicular alliti'de, not according to the quantity of their 

 matter; whence the fame weight of the atmofphere fup- 

 ports the quickfilver that fills the tube A C D, and the 

 ciftern B, as would fupoit the mercury in the tube 

 alone. 



This laft, however, with its excellencies, has great defects : 

 for, by reafon of the attraction between the parts of the 

 glafs and of the mercury (which Dr. Jurin has (hewn to be 

 confiderable), witli the length of the fcale (confequcntly 

 the quantity of motion), and the attrition againft its fides, 

 efpecially in fudden rifes and dtfcents, the mercury breaks, 

 fome parts of it are left behind, and the equability of its 

 rite and fall ruined. Some therefore prefer the 



Inclined barometer, or diagonal, of fir Samuel More- 

 land, where the fpace of variation is confiderably larger 

 than in the common one, and yet the rife and fall 

 more regular than in the others — Its foundation is this; 

 that in a Torricellian tube BC (Jig. 8i.) inchncd at any 

 angle to the horizon, the cylinder of mercury equivalent to 

 the weight of the atmofphere, is to a cyhnder of mercury 

 equivalent to the fame in a vertical tube, as the length of 

 the tube BC to the perpendicular height DC. 



Hence, if the height DC be fubtriple, fubquadruple, 

 &c. of the length of the tube, the changes in the diagonal 

 barometer will be triple or quadruple, &c. of the changes 

 in the common barometer. This barometer will fcarce 

 allow its tube to be inclined to the horizon at a lefs angle 

 than 45°, without undergoing the inconveniency of the hori- 

 zontal one. 



Mr. Orme, in order to obviate fome of the objections to 

 which the diagonal conftru£tion of the barometer is liable, 

 purified the quickfilver from its drofs and earthy particles 

 by diftillation ; and when the tube was filled with a cer- 

 tain quantity of mercury, difcharged the remaining air by 

 an intenfe heat fufRcient to make the mercury boil ; and 

 he continued this operation for four hours. In the pro- 

 cefs, an innumerable quantity of fmall particles were emit- 

 ted, and when no more bubbles rofe in the tube, the mer- 

 cury appeared extremely bright, but funk lower in the tube 

 than when it was firft put in, by two inches. Phil. Tranf. 

 Abr. vol. viii. p. 455. 



The -wheel barometer was a contrivance of Dr. Hooke, in 

 1668, to make the alterations in the air more fenfible ; the 

 foundation of this is the common vertical barometer, with a 

 large ball above, and turned up at the lower end, with the 

 addition of a couple of weights A and B (Jig. 82.) hanging 

 on a pulley, the one of them playing at hberty in the air, 

 the other refting on the furface of the mercury in the in- 

 verted tube, and rifing and falling with it. 



Thus is the motion of the mercury communicated, by 

 means of the pulley, to an index which turns round a gra- 

 duated circle ; and thus the three inches of vertical afcent 

 are here improved to five, fix, or more, at pleafure. 



But the friftion of the axis of this index, and more 

 efpecially when it has contracted fome ruft, generally ren- 

 ders this fort of barometer ufelefs ; and, at beft, the gra- 

 duation of inches on the circle can only be confidered 

 as a fcale of motions of the mercury in its tube; for the great 

 variation of the height of tlie furt.-tce of the mercui-j- in the 

 tube below will perpetually falfify the inches and tenths 

 upon the plate above. In a juft or ftandard barometer, the 

 inferior furface of the mercury in the ciftern or tube below 

 ihould either be invariable, or reducible by a prtfiing fcrew 

 to a fixed or determinate gauge point. 



The wheel barometer has lately been obtruded upon the 

 public by the llroUing Itahan hawkers in our ftreets ; but 

 the imperfett manner in which thefe barometers are con- 



ilruded. 



