THERMOMETER. 



calil)- penetrate* and reaches the uimoft paruof the cylindric 

 bulb, (lid caufts tlio whole content to expand unlfonnly, 

 aud the mercury to rife alraoll immediately; whereas in 

 ihennonieters with a fplierical bulb it is fecn iirft to fall, 

 and then to rife. Tins phenomenon has been long fince 

 noticed both in Florentine and mercurial thermometers, 

 wlif n they are fuddcniy plunged into a healed liquor, the 

 fpirit of wine or mercury tirll defcends, and then afcends ; 

 aiid when they are plunged into a cold fluid, the included 

 hquor lirll afcends and then defcends : this is the more re- 

 markable in thermometers whofe bulb is made of thick 

 elafi ; and the reafon of the phenomenon is obvious. The 

 bulb of glafs is fooner afTeded by the heat or cold applied 

 to it than the included fluid ; and as tlie glafs expands by 

 heat, the capacity of the bulb is enlarged, and the hquor 

 defcends in the tube, but being condenfed by cold, and its 

 capacity diminifhcd, the liquor is prefled upwards in the 

 tube : and both thefe efFefts continue till the heat and cold 

 equally afleft the inclofed fluid. Hence it follows, that all 

 the variations of afcent and defcent, to which the fpirit or 

 mercury is fubjeft in the thermometer, are only the differ- 

 ence of the rarefaiftions and condenfations of glafs, and of 

 the contained fluid. Hift. Ac. Royal, 1705. 



The cone, or cyhnder, of the thermometer is made of 

 glafs of a moderate thicknefs, left, when the exhaufted tube 

 IS hermetically fealed, its internal capacity fliould be di- 

 miniftied by the weight of the ambient atmofphere. When 

 the mercury is thoroughly purged of its air and moifture 

 by boihng, the thermometer is filled with a fuflicient quan- 

 tity of it ; and before the tube is hermetically fealed, the 

 air is wholly expelled by heating the mercury, fo that it may 

 be rarefied and afccnd to the top of the tube. To the fide 

 of the tube is annexed a fcale {ice Jig. 4.) which Fahrenheit 

 divided into fix hundred parts, beginning with that of the 

 fevere cold which he had obferved in Iceland in 1709, or 

 that produced by furrounding the bulb of the thermometer 

 with a mixture of fnow or beaten ice and fal ammoniac or 

 fca-falt. This he apprehended to be the greatefl degree of 

 cold, and accordingly he marked this, as the beginning of 

 his fcale, with o ; the point at which mercury begins to 

 boil, he conceived to fliew the greateft degree of heat, and 

 this he made the limit of his fcale. The diftance between 

 thefe two points, he divided into fix hundred equal parts 

 or degrees ; of which 32 reckoned from o, indicates the 

 degree of cold when fnow or ice thaws naturally, or water 

 begins to freeze, and this is called t\ie freezing point : and he 

 marked the heat of boihng water with 212, &c. In order 

 more particularly to explain the divifions of this fcale, and 

 to (hew how the dilatation and condenfation of the mercury 

 are eftimated by it, we may obferve that the bulb is fup- 

 pofed to contain, according to Boerhaave and Mufchenbroek, 

 1 1 1 24 parts of quickfilver, which ftands at the lowefl; mark, 

 or gr. o, in an intenfe cold, &c. as above determined : if the 

 bulb be immerged in fnow or ice thawing naturally, or in 

 water beginning to freeze, the quickfilver is dilated, and 

 rifes in the tube 32 of thefe 11 124 parts; and therefore 

 the fpace of the tube from gr. o to the freezing point gr. 

 32, IS divided into thirty-two equal parts. When the ther- 

 mometer is placed in water brought to a ftrong boiling at a 

 middle flate of the atmofphere in places near the level of the 

 fca, when the mercury in the barometer ftands at about 30 

 inches or a very httle under it, the quickfilver is dilated 212 

 of thefe parts beyond its original bulk of 1 1 124, fo as now 

 to poffefs m the bulb and tube together a fpace equal to 

 1 1336 fuch parts ; and the fpace from gr. 32 to gr. Z12, is 

 divided into 180 equal parts or degrees of the thermometer; 



which, if the tube te long enough, may be protraftifd as far 

 as is convenient. It may extend well enough to gr. 600, 

 and not much farther, for with a heat but little greater than 

 that the mercury begins to boil. 



Dr. Boerhaave, in one place, makes the number of parts 

 into which the mercury in the bulb is fuppofed to be divided 

 to be 10782 inftead of 11 124, and in another place ftates it 

 at 11520, which Dr. Martine apprehends to be nearer the 

 truth, or about 1 1790 parts ; and he thinks the eafieft and 

 fureft method is to fill the bulb and tube, without being 

 folicitous about the bulk of the quickfilver, fo tliat in freez- 

 ing water, or melting ice, the mercury (hall ftand at a con- 

 venient height, which muft be very nicely marked gr. 32 ; 

 and then as accurately to obferve where it ftands when 

 dilated by the heat of boihng water to ^r. 212. The inter- 

 mediate fpace is then divided into 1 80 degrees, which fcale 

 may be protrafted upwards or downwards as far as we (hall 

 judge convenient. See Fixed Points 0/" Thermometers. 



In the above method of graduating the fcale, the bore of 

 the tube is fuppofed to be perfeftly cylindric, which cannot 

 always be obtained. But though it be tapering or fome- 

 what unequal, it is eafy to manage that matter, in the manner 

 propofed by the abbe Nollet (Lemons dc Phyf. Exp. 

 torn. iv. p. 376.) by making a fmall portion of the quick- 

 filver, e.g. as much as fills up an inch or half an inch, Aide 

 backward and forward in the tube ; and thus to find the 

 proportions of all its inequalities, and from thence to adjull 

 the divifions to a fcale of the moft perfeft equahty. See 

 Obfervations on the ConJlruSion q/" Thermometers. 



Other thermometers of a fimilar conftruftion have been 

 accommodated to common ufe, the fcale of which is only a 

 part of that above defcribed. They have been made of a 

 fmall fize and portable form, and the tube with its annexed 

 fcale has been enclofed in another thicker glafs hermetically 

 fealed, in order to prelerve it from injury. Mr. Ramfden, 

 at the defire and for the ufe of Mr. Hunter in his experi- 

 ments on the heat of animals and vegetables, conftrufted 

 very fmall thermometers, fix or feven inches long, and not 

 above two twelfths of an inch thick in the ftem ; having the 

 external diameter very little larger than that of the ftem, on 

 which was marked the freezing point. The ftem was em- 

 braced by a fmall ivory fcale, fo as to Aide upon it eafily, 

 and retain any pofition. Upon the hollow furface of this 

 fcale were marked the degrees which were feen through the 

 ftem. Phil. Tranf. vol. Ixviii. parti, p. 8. 



Several varieties of thermometers are conftrufted for phi- 

 lofophical purpofes. For comprehending the whole range 

 of thermometrical temperature from the moft intenfe artifi- 

 cial cold to the boihng point of mercury, it is necelTary to 

 be provided with a very long tube ; but for moft chemical 

 purpofes, it need only be graduated to about ten degrees 

 above the boihng point, which will reach the temperature of 

 moft faline folutions when boihng. For experiments in in- 

 tenfe cold, a fpirit thermometer fhould be graduated about 

 100 degrees below o, and the lower extremity of the fcale 

 (hould be at feme diftance from the bulb, that the tempera- 

 ture may be obferved without lifting the bulb out of any 

 deep veflel that may contain the freezing mixture. The 

 moft delicate and fenfible thermometers are made with a very 

 fmall bulb, fcarcely larger than the ftem, and a tube of an 

 extremely narrow bore, not larger than a horfe-hair. For 

 chemical purpofes alfo, the fcale (hould either be fcratched 

 on the glafs itfelf, or, as this is difficult to be feen in a com- 

 mon light, an ivory fcale (hould be attached without reach- 

 ing fo low as the bulb, that the latter may be fafely im- 

 merfed in acid or corrofive hquors. 



In 



