B A R 



B 



R 



tionof the ciilern, Sec. of the portable barometer, according 

 to the conftnictjon of Mr. Hawcs, lately an eminent iniliu- 

 inent-maker in London. A feftion of the cillcm h repre- 

 fented in P/aiiXll. of P nenmatks , f.g. 102. AAAABB 

 is the cillcrn : the part AAA A, which contains the quick- 

 filver, is madj of wood, with a bottom of leather C, glued 

 on the wooden ring DD, and preffed clofe agaiiilt the 

 wooden cylinder by means of the fcrew at EE, which fcrews 

 on the brafs cover or collar FF that covers the cylindric 

 ciiterii AA. This collar has a ilep at the top, as leen at 

 G, to prevent its (lipping, while it prcfTcs the ring DD againll 

 the wooden cylinder AA. When fcrewcd tight, the quick- 

 filver HH is prevented from efcaping. II is part of the 

 tube of the barometer, drawn nearly to a point, and covered 

 with an ivory cap KK for defending it againft injury. L.L 

 is a fcrew with a broad circular top Q^ by means of which 

 the leather C forces up the quickfilver ib as to fill the tube, 

 when the inllrument is carried from one place to another. 

 In order to prevent the ofcillations of the quickfilver from 

 breaking the tube by hidden jerk^^, a pin a with a head b 

 pafTes through the fcrew LL, ; this pin has on the under 

 fide of the head a fpiral fpring to counteract the violence 

 of a fudden motion. The two nuts M, N, areuled to raife 

 or deprefs the fcrew LL, and confcquently the quickfilver ; 

 the proper height of which is indicated by the floating gage 

 00, the top of whofe ftem P correfponds to the top and 

 outilde of the ciilern. When the barometer is not in ufe, 

 the gage and aperture are covered by the plate e, which 

 effecluaUy confines the quickfilver, the under fide of e being 

 covered with leather. The lower end of the fcrew LL is 

 flit up as high as c, and carries a crofs pin d pafung througli 

 the bottom of the pin a h \.qi prevent it from rifing too high. 

 F'lg. 103. reprefents a fquare frame to be fcrewed on the 

 part YiSiJig. 103, and connected by wires from the angles 

 to the legs as feen in the perfpcftive view in jig. loi. This 

 is ufed to prevent the barometer from vibrating. 



The nonius is exhibited at large in _;^. 104. A is a 

 fcrew with a milled head tapped into the piece B, and alfo 

 let into and moveable in the piece C in the manner repre- 

 fented at D in Jig. 105. which is a fide view. 



B and C in Jig. 106. are horizontal feftions of B and C, 

 Jig. 104. The fpring a of the piece B is confiderably 

 ilronger than that of C ; fo that it requires much greater 

 force to make it Aide up and down, whilft C, which Aides 

 very eafilv, is moved by turning the milled head E ; and 

 thus the lower furface of C is made to coincide with the 

 upper furface of the mercury at F ; and, befides, both the 

 piece B and the nonius C may be depreffed or raifed at 

 pleafure, as occafion requires, for a due adjuftment of the 

 nonius. Behind the plate a b, in the perfpcttive view 

 Jig. loi. hangs a pendulum fufpinded at the point a which 

 ferves for fetting the inilrument vertical ; and when it is 

 brought into this pofition, a mark on the bob coincides with 

 another on the plate, as feen at b. When the inilrument 

 is not in ufe, a fork connefted with the fcrew c is puQied up, 

 and prevents the pendulum from, (haking. 



In order to adapt the portable barometer more completely 

 to the purpofe of meafuring heights, in which ufe of it 

 pecuhar accuracy of obfervation is neceffary, it (hould be 

 fuiTiilhed with two microfcopes or magnifying glafles, one 

 of which fhould be placed at the beginning of the fcale ; 

 and either this (hould be moveable, fo that it m.ay always be 

 brought to the furface of the mercury in the ciilern, or the 

 ciftern (hould be fo contrived that its furface may always be 

 brought to the beginning of the fcale. By this glafs the coin- 

 cidence may be accurately perceived. The other microfcop'e 

 luuft be moveable, fo as Co be fet oppofite to the furface of 



the mercury in the tube ; and the fcale (liould be fumifhed 

 with a vernier, which diviJ/s an inch into icoo p'rts, and 

 coiiilrufted of matenaU, the expar.iion of which is precifely 

 afcertained. For an account of many ingenious contrivances 

 to make the barometer accurate, portable, and commodious, 

 the reader may conlult jVIagcllan's '• DilT. de Diverfes Inftr. 

 de Phvi." Phil. Tranf. vol. Ixvii. vol. Ixviii. Journ. de 

 Phyf. xvi. •^92. xviii. 391. xix. 308. 34.6. xxi. 436. xxii. 390. 

 Sulzer, Act. Hclvet. iii. 259. De Luc, Rcchsrchcs, &c. 

 ubi fupra. Cardin. de Luynes, Mem. Par. 1768. Van 

 Swinden's Pofitiones Phyficae. Com. Acad. Petrop. i. Id. 

 Nov. ii. 200. viii. 



Mr. Magtiian, in his edition of " Cronfttdt's Mi- 

 neralogy," has fliewn that great errors may arife in barome- 

 trical meafureraents for want of due attention to the fpecific 

 gravity of the mercury with which barometers are filled. 

 If two barometers, each 30 inches high, and in every other 

 refpecl fimilarly circumftanced, be tilled with mercury of 

 diflercnt fpecific gravity, that of the one being 13,62, and 

 that of the other 13,45, the error in the rciult would be 

 no lefs than 327 feet ; becaufe the heights of the mercurial 

 columns in each barometer muft be in the inverfe ratio of 

 their fpecific gravities; viz. 13,45 '• 'S'^^ " 3° '• 33>379- 

 But the logarithm of 30 is 4771,21, and that of 30,379 is 

 4825,73, neglefting the indices, and their difference is 54,52, 

 which (hews that there is a diiterer.ce of 54,52 fathoms 

 or 327 feet in the altitudes of the two places, where the 

 barometers (hoidd have been ftationed, though in reality 

 they were on the fame level. But if the fpecific gravity of 

 the mercury in the two barometers were according to the 

 different (tatements of Bergm.an and Fourcroy, the one 

 14,1 10 and the other 13,000, (and this may happen to be 

 the cafe, as the heavieft. is commonly reputed to be the pureft 

 mercury.) the error muft have amounted to 355,76 toifes, 

 or 21344 feet, becaufe 13,000: 14,110 :: 30 : 32,561. 

 But the logarithm of 30 is 4771,21, and that of 32,561 is 

 5126,97, and the difference, or 355,76, (hews that the 

 error (hould amount to fo many fathoms, or 2134,5 feet. 

 See the fequel of this article. 



Barometer, Phenomena oj the. Thefe are the variations 

 of height in its mercurial column, for afcertaining which 

 many contrivances in the ftrufture of the barometer have 

 been propoled ; the principal of which have been delailed 

 in the preceding articles ; and the fubjeft will be further 

 purfued in the fequel. The ufes to which thefe phenomena 

 have been fubiervient, are the prediction of tlie weather 

 from the variable weight of the atmofphere, indicated by 

 the rife and fall of the mercury in the barometer, and the 

 meafurement of altitudes, to which they have been lately 

 appUed with lingular affiduity and fuccefs. 



The phenomena of the barometer, confidered as a " wea- 

 ther glafs," have been very differently dated and explained 

 by various writers ; and they are fo precarious, that it is 

 extremely difficult to form any fixed and general rules con- 

 cerning them. Although we have reafon to beheve, that 

 the barometer never fails to indicate a ftorra, or any very 

 great change of weather, for feme hours before it occurs ; 

 yet its variations afford no indications or prognoftics that 

 are abfolutely certain, with refpecl to thofe lefs confiderablc 

 changes to which the weather is fubjecl in our variable 

 climate. With certain reftriAions, they afford fome ground 

 for probable coiijefture ; and thefe reftriftions are to 

 be determined merely by the fagacity cf long-continued 

 obfervation and experience. Striilly fpeaking, the height 

 of the mercury in the barometer hath no immediate and 

 neceffary connexion either with i-ain or fair weather. That 

 its variable height is the immediate confeq^ueuce of the 



variable 



