4^0 Ohscrvatmn respecting a New Scale 



not that the human body, from a well-known provision iii 

 the animal oeconomy, has the innate power of resisting, or 

 rather of diminishing, the effect of extreme heat to a cer- 

 tain extent. 



My chief object in the present paper has been to inforte 

 the propriety of the alteration suggested, in the place of 

 •the zero in a thermometrical scale. With respect to the 

 fittest mode of division to be adopted, I have preferred 

 that of Fahrenheit, for the reason above stated ; but since 

 the mode of decimal divisions seems at the present time 

 to be gaining ground, I have exhibited a scale of the centi- 

 grade division likewise in the annexed table. (SeeTab. H.) 



It will be apparent, in adopting the centigrade divi- 

 sion in the stead of Fahrenheit, that there will be a dif- 

 ference, occasionally, amounting to -^dof a degFee, between 

 them ; bnt as the three leading points correspond precisely, 

 so slight a variation in the others is of little or no conse- 

 quence; and indeed all the points cannot be perfectly ac- 

 curate upon any uniform scale : thus, the precise point at 

 which quicksilver freezes was ascertained to be — 38 and 

 f- of Fahrenheit, but is marked at —39. 



There are fouf different proportions of the scale, which 

 I should recommend to be used according to circumstances ; 

 VIZ. Igt, a portable scale extending from —30 to +30 ; 

 fidly, oi^ for ordimiry meteorological use in this climate, 

 extending from ~ 70 to -|-70*; 3dly, from —100 to 

 + 150 1 ; and 4thly, a thermometer of coloured alcohol, ex- 

 tending from — 162 to -I- 40 :J. 



I flatter myself that I have adduced sufficiently satisfac- 

 tory reasons for the adoption of the scale I have here sug- 

 gested; and, when the usual prejudices for an old sy- 

 stem have subsided, and the apparently ins«rmountabIe 

 difficulties attending the introduction of a new one are con- 

 quered, that the one I have now offered to consideration 

 will be admitted to be perfectly reconcileable in a philoso- 

 phical view; convenient and useful in application; not 

 liable to be affected by new discoveries or fresh theories in 

 future; and, in short, a scale dictated as it were by nature, 

 for universal application. 



* A local scale of this kind in every climate w6ii!d he particularly useful 

 in the place itself, and iuterestin|; every where else. 



+ The freezing point of quickijlver upon this scale is — lOl ; but as quick- 

 silver ceases to measure temperature at — 100, the scale of a quicksilver 

 fhermometer may terminate there. 



I Quicksilver is preferable to alcohol for thermometers, as far as Its 

 range goes, for several reasons ; but, principally, because it is found to be 

 siOTc UDifonn in its expaosion by htat than aU(^l or spirit. 



P.S.~Be- 



