BAR 



variable prefTure of the atmofphei-e, is a facl thai admite of 

 no doubt ; but the caofca of this variable preffure have not 

 yet been fully and fatisfaftorily afcertained ; and how far 

 the ftate of the weather, in all itJ minute and fiiddcn changeH, 

 depends upon it, is a queflioii that Hill remains to be deter- 

 mined. M. Pafcal was one of the lirfl perfuns who partieu- 

 larly obferved the variations of the barometer, and referred 

 them to correfponding ehanges in the weight of the air ; 

 but he acknowledges, that it is very d'.fficult to explain 

 both the one and the other, as well as the conneftion that 

 fubfifts between them. He obfervcs, in general, that the 

 mercury is commonly higheft in winter and loweft in fum- 

 mer ; that it is leaft variable at the foliliccs, and motl va- 

 riable at the equinoxes : and he add?, in direft contradidian 

 to later experience, th;\t the mercury ufually falls in fine 

 weather, and that it rifcs when the weather becomes cold 

 or the air is loaded with vapours. M. Pafcal was followed 

 by Perrier, Deal, Wallis, Oarcin, Garden, Lifter, Halley, 

 Gerften, De la Hire, Mariotte, I.e Cat, Woodward, 

 I-cibnitz, De Mairan, Bernouilli, Mufchenbroek, &c. ; all 

 of whom have given diflereut foluiiona of the phenomena 

 of the barometer. 



The principal obfervations, that have been made on the 

 variations of this inllrument, are fummed up by Mr. Kirwan 

 (Irifh Tranf. vol. ii. p. 4ft, 5:c.) in the following particu- 

 lars. 



I. The more confiderable elevations and depreffions of 

 the mercnr)- in the barometer happen at a very fliort inter- 

 val of time in places very remote from each other. This 

 eonefnondence was obferved by Mr. Derham in 1699 be- 

 tween the heights of the mercury at Upminfter in EfTex 

 and Townley in Lancafliire ; and afterwards by Mr. Ma- 

 ralJi between the variations at Paris and Genoa, at 

 the dillance of nearly four degrees of latitude, who 

 adds, during thefe variations different winds prevailed at 

 thefe places. But Mr. Kirwan obferves,that where there is a 

 confiderable difference of longitude, the like agreement is 

 not fou:id. 



H. The deviations of the mercury from its mean annual 

 altitude arc far mare frequent and extenfive in the neigh- 

 bourhood of the poles than in that of the equator. At 

 Peterdjurgh, in 1725, the mercury once ftood at the ftupen- 

 dous height of 31,59 inches, if we may credit Mr. Confett ; 

 and yet it has been leen fo low as 28, 14 inches. In the nor- 

 thern parts of France the variations are greater than in the 

 fouthern : at Naples they fcarcely exceed one inch. In Peru, 

 under the equator, and at the level of the fca, they amount 

 only to two or three tenths of an inch ; but in other parts, 

 within a few degrees of the line, on the approach of the 

 rainy feafon or of hurricanes, the barometer fiRlIs an inch or 

 more. 



HI. The variations without the tropics arc greater and 

 more frequent in the winter than in the fummer months. 



IV. The variations are confiderably fmallei' in very ele- 

 vated fituations than on the level of the fea. Tiius M. Bou- 

 guer obferved, that on the coafts of Peru the variations ex- 

 tended to i ofanhich: at Quito, elevated 93 74 feet above the 

 fea, they comprehendt-d only 0,083 "^ '"' inch. M. Sauffure 

 made fimilar obfervations in Savoy, as did Mr. Lambert in 

 SwifTerland. 



V. The mean height of the barometer on the level of the 

 fea in moll parts of the globe hitherto examined, is about 

 30 inches. M. Bouguer, under the line, obferved it at 

 29,908 inches; but as his barometer was not purged of air 

 by fire, it llood lower than it fhould have done. Sir George 

 Shuckburgh (Phil. Tranf. vol. Ixvii. p. 586.), on a mean of 

 federal obfervations on the coalls of Italy and England, 



BAR 



found it at 30,04, when the temperature of the mcrciiry WSI ' 

 55'^, and that of the air 62°. The mean beight of the ba- 

 rometer in London, upon an average of two obfervations in 

 every day of the year, kept at the houfe of the Royal So* 

 ciety, for many years pall, is 29,88 ; the mean temperature 

 or height of the thermometer, according to the fame, being 

 58°. The greated height obferved by fir G. Shuckburgh, 

 Dec. 26, 1778, in London, was 30,948 inches, the thermo* 

 meter being at 47° ; and reduced to the heat of 50°, it vi'as 

 30,957: and this, he fays (Phil. Tranf. vol. Ixix. p. 370.), is 

 the greated height, which, as far as he has been able to coi- 

 led, it has ever been feen to (land at in any country, where 

 obfervations have been made and recorded, fince the firll 

 invention of this inftrument. In the pi-oximity of the poles, 

 fays Mr. Kirwan, the annual mean heights of the barometer 

 differ m.uch more from the above llandard than in the more 

 fouthern parts of our hemifphere. 



In eftimating the conuetlion of the variations of the 

 barometer with the weather, Dr. Halley has propofed the 

 following rule;^ : 



I. In calm weather, when the air is inclined to rain, the 

 mercury is commonly low. 



II. In fercne and fettled weather, and alfo in calm and 

 frofty weather, the mercury is generally high. 



III. Upon very liigh winds, though not accompanied 

 with rain, the mercury finks loweft, regard being had to the 

 point of the compafs from which the wind blows. 



IV. The greatell heights of the mercuiy are found upoii 

 cafterly and north-eallerly winds, other circumllances being 

 alike: to which it may be added, that under a foutherly 

 wind it is commonly low. The above four obfervations, 

 made by Dr. Halley in England, feem to be moll univerfalj 

 as they were found by Mr. Mclander (Schwed. Abhand" 

 1773, S. 2t5) to apply to lat. 39°, and by M. de Luc to 

 lat, 46". 



V. After very great ftorms of wind, when the mercury 

 has been very low, it generally rifes again very fad. 



VI. The more northerly places have greater alterationj 

 of the barometer than the more foutherly. 



VII. Within the tropics, and near them, there is little of 

 no variation of the mercury in all weatlicrs. At St. Helena 

 it is little or nothing; at Jamaica T^jjths of an inch; whereas 

 in England it amounts to Zj inches, and at Peterfijurgh to 

 3 I nearly. 



Dr. Beat, who adopted the opinion of M. Pafcal, obferves, 

 that, citleris paribus, the mercury is higher in cold weather thali 

 in warm ; and ufually in the morning and evening hirrherthan 

 at mid-day : that, in fettled and fair weather, it is hio-her 

 than either a little before or after or in the rain; and that 

 it generally defcends lower after rain than it was before it. 

 And he afcribcs thefe effctls to the vnponrs with which the 

 air is charged in the former cafe, and which are difperfed by 

 tlie falling rain in the latter. If it chance to rife higher 

 after rain, it is generally followed by a fettled fertnity. 

 He adds, that there are frequently gi-eat changes in the air, 

 without any fettled alteration in the baiomcter. 



An ingenious author obferves, in relation to -this ufe of 

 barometers, that, by their means, we may regain t!ie know- 

 ledge, which dill refides in brutes, and which we have for- 

 feited'by not continuing in the open air as they generally 

 do, and by our intemperance corrupting the crafis of our 

 organs of fenfe. 



Mr. Patrick's rules for judging of the weather by the rife 

 and fall of the mcrc\n-y in the barometer, have been much 

 approved, and are to be accounted for on tlie fame principles 

 with thofe of Dr. Halley. They are as follow : — i. The 

 rifing of the mercury prefages, in general, fair weather; 



and 



