THERMOMETER. 



Til 1733, M. dcl'ir,.- of PetJrd'jrT coiiilruifted a imrcu- 

 ».W thermometer [kcfig. 4.) on the principles of Reaumur's 

 fpirit thermometer, in his thermometer, the whole bulk of 

 quickfilver, when immcrged in boiling water, is conceived to 

 be divided into 10,000, or rather 100,000 parts ; and from 

 this one fixed point, the various degrees of heat, cither above 

 or below it, are marked in thefe parts on the tube or fcale, 

 by the various ex panfion or contraftion of the quickfilver in 

 all the imaginable variety of heat. Dr. Marline apprehends 

 it would have been better if M. de I'KIc had made the in- 

 teger of 100,000 parts, or fixed point at freezing water, and 

 from thence computed the dilatations or condenfations of the 

 quickfilver in thofe parts. All the common obfervations 

 of the weather, &c. would have been expreffed by numbers 

 increafing as the heat incieafed, which is the more natural 

 way ; nor would there have been any great incongruity, or 

 inconvenience, in exprefiing, after the manner of Reaumur, 

 the few obfervations that occur below fimple freezing by 

 numbers of contraftion below gr. O, or i0O,00O. How- 

 ever, in praftice, it will not be very eafy to determine ex- 

 adly all the divifions from the alteration of the bulk of the 

 contained fluid. And befides, as glafs itfelf is dilated by 

 heat, though in a lefs proportion than quickfilver, fo that it 

 is only the excefs of the dilatation of the contained fluid 

 above that of the glafs that is obfervable ; if difl"erent kinds 

 of glafs be differently affefted by a given degree of heat, 

 this will make a feeming difference in the dilatation of the 

 quickfilver in the thermometers, conftrufted ie the New- 

 tonian method, either by M. de Reaumur's or M. de 1 !""'« 

 rules. Accordingly, it has been found, that the quickfilver 

 in thermometers, conftrufted in M. de I'lde's way, has ftood 

 at different degrees of the fcale when immerged in thawing 

 Inow. In fome it was at_fr. 154, in others at 156, and in 

 another at 158 : and it appears by M. de I'Ifle's own ac- 

 count, that his thermometers difagree confiderably from one 

 another. Celjlus's thermometer has been chiefly ufed in 

 Sweden, and hence it has been called the Siucdtfo thermo- 

 meter. The French chemifts have lately adopted it, under 

 the name of centigrade thermometer. See Fixed Points of 

 Thermometers, and the table at the clofe of this article. 



Thermometer, Metalline, is a name given to a machine 

 compofed of two metals, which, whilft it indicates the va- 

 riations of cold and heat, ferves to correft the errors that 

 refult from hence in the conftruftion of pendulum clocks. 

 Inftruments of this kind have been contrived by Graham, 

 Le Roy, in 1738, EUicot, Harrifon, &;c. See Compound 

 Pendulum. 



We have alfo an account of inftruments of this kind in- 

 vented by Mortimer, Frotheiingham, and Fitzgerald, in the 

 Phil. Tranf. vol. xliv. p. 689, vol. xlv. p. 129, and vol. li. 

 p. 82J ; to which we muft refer for a particular defcriptioa 

 of each, illuftrated by figures. 



M. de Luc has likewife defcribed two thermometers of 

 metal, which he ufes for correfting the effefts of heat upon 

 a barometer, and an hygrometer of his conftruftion con- 

 nefted with them. In one of thefe, a ilrong rod of well- 

 hardened brafs, fupports upon a^ edge, at a convenient 

 diftance from the centre of motion, a lever, vvliich holds the 

 fcale of the barometer fufpended, and makes it rife or fall 

 by the dilatation or condenfation of the brafs rod, as the 

 quickfilver rifes or falls in the barometer, by the conefpond- 

 ing variations of heat. This fcale of the barometer, when 

 it moves, draws or loofens a thread of filk-grafs, which 

 goes over a fmall pulley placed upon the fame axis with a 

 much larger one, to which the tcale of the hygrometer is 

 hung likewife by a fimilar thread, which thus varies, by the 

 proportion of the diameters of the pulleys, as the heat makes 



Vol. XXXV. 



the quickfilver in the hygrometer vary. This inftrlinient is 

 convenient for meteorological obfervations : becaufe it faves 

 one obfervation and two corrcftions for the heat ; but it is 

 nccefl'ary from time to time to correft an irregularity in it, 

 which is eafily perceived by means of an index, carried by 

 the moveable fcales of the two inftruments, which, going 

 over immoveable fcales of the fame fort, ftiews their dtfTcr- 

 ence of height. When this difference is no longer con- 

 formable to the indication of the thermometer, it is eafily 

 reftified by turning fmall pegs, on which is twifted the 

 thread of filk-grafs, which ferves for the fufpcnfion of the 

 fcales. The irregularity juft mentioned confilts in this, that 

 when the heat, after having varied, returns to the fame point 

 of the quickfilver thermometer, the metallic thermometer 

 does not return to it exaftly, but varies nearly in the fol- 

 lowing manner ; during the fummer, the latter gains coii- 

 ftantly on the former, i. e. amidft its variations, it always 

 preferves a fmall part of the lengthening, which is at that 

 time its ordinary ftate. In winter, on the contrary, it be- 

 comes infenfibly ahttle too fhort. The other metallic ther- 

 mometer, which is more curious than ufeful, on account of 

 its greater irregularity, confifts of a rod of lead, which, 

 communicating by a thread of fdk-grafs with a fmall pulley 

 fixed to the fame axis with a greater one, condufts, by mear.s 

 of another pulley, a needle through whofe axis, which is 

 bored, pafTes another axis that carries the needh of a pulley 

 barometer. Thus this inftrument marks the heat and 

 weight of the air upon two concentric circles, by means of 

 two needles turning upon the fame centre, as in clocks ; be- 

 hu.' vvhich, the needle of the thermometer points out upon 

 a thira v'""le the correftion for the heat, to be made on the 

 barometer. See Phil. Tranf. vol. Ixviii. part i. p. 43 7, &c. 



Thermometers, Oil. To this clafs belongs fir Ifaac 

 Newton's thermometer, conftrufted in 1701, with linfeed 

 oil inftead of fpirit of wine. This liquor has the advantage 

 of being fufficicntly homogeneous, and capable of a con- 

 fiderable rarefaftioii, not lefs than fifteen times greater than 

 that of fpirit of wine. It has not been obferved to freeze 

 even in very great colds, and it is able to bear a great heat, 

 about four times that of water, without boihng. With thefe 

 advantages it was made ufe of by fir Ifaac Newton, who 

 difcovered by it the comparative degree of heat for boiling 

 water, melting wax, boiling fpirit of wine, and melting tin ; 

 beyond vvhich it does not appear that this thermometer was 

 applied. The method he ufed for adjufting the fcale of this 

 oil-thermometer was as follows : fuppofmg the bulb, when 

 immerged in thawing fnow, to contain 10,000 parts, he found 

 the oil expanded by the heat of the human body fo as tq 

 take up one thirty-ninth more fpace, or 10256 fuch parts ; 

 and by the heat of water boihng ftrongly, 10725 ; and by 

 the heat of melting tin, 11516. So that, reckoning the 

 freezing point as a common Hmit between Iieat and cold, he 

 began his fcale tliere, marking it gr. o, and the heat of the 

 human body he made^r. 12 ; and, confequently, the degrees 

 of heat being proportional to the degrees of rarefadtion, or 

 10256: 10725, i. e. 256 : 725 :: 12 : 34, the heat of boil- 



725 X 12 



ing water was expreffed by gr. 34 ■ 



256 



and that 



of the melting tin by gr. 72. Phil. Tranf. No. 270, or 

 Abr. vol. iv. part ii. p. 3. 



Although in this graduation fir Ifaac Newton does not 

 fpecify any degree of cold below that of freezing water, yet 

 it would be eafy to protraft his fcale downward below^r. o, 

 or the freezing point, and thus to adapt it for eftiraating 

 greater degrees of cold, like other thermometers. But 

 there is another infuperable inconvenience that attends all 

 3 S thcr- 



