348 Obfervations and Experiments on Light and Heat. 



tometer was placed beyond the fpeftrum, though it approached the margin, whether 

 above, below, or at the fides, no efFe£l whatever was perceived. And as that inftrument 

 greatly furpaffed, in feiifibility and regularity of performance, the fineft thermometer, it is 

 only what might be expected, that I could not perceive any alteration, if I employed ther- 

 inometers of uncommon delicacy and of various conftruflions, with black or pellucid 

 bulbs, having their fcales Subdivided into tenths of a degree, and capable of indicating ftill 

 fmaller differences. There is a circumftance, however, which being overlooked, might 

 lead to egregious errors. If the fpe£lrum be received on the (land, the inftrument brought 

 near it will be very fenfibly afFe£i:ed5 owing partly to the light reflefled, but chiefly to the 

 aftion of the heated air accumulated over the illuminated furface. In fliort, the operations 

 are deranged in every cafe where the inftrument is not completely infulated, and is remote 

 from all folid materials which might detain the light. And it is worthy of remark, that 

 the afflux of light muft raife the temperature of a flat furface of imperfedt condu£ling fub- 

 ftance, fuch as wood or pafteboard, more than four times as much as it will the blackened 

 ball of a thermometer; for, in the former cafe, the heating and cooling caufes are exerted 

 on the fame fpot, but in the latter, the quantity of light or heat received is only as 

 the central feftion of the ball, while the whole furface, which is quadruple that fpace, is 

 expofed to the cooling influence of the ambient air. There are other circumftances, too, 

 which tends to augment that difference of effeft, but which I fhall not ftop to mention. 

 On the whole it is mod difficult, where an a£live fource of heat exifts, to obtain an uni- 

 form temperature, particularly in a fmall room ; and I might point out caufes, not yet ap- 

 prehended, which will notably afFe£t a thermometer. 



Thefe hints will ferve as a clue for dete£ling Dr. Herfchel's miftakes. The firft 

 circumftance that begets fufpicion is the large quantities marked by his thermometers, 

 and which are not much inferior to the full force of the fun-beam. To procure the colours 

 of the fpe£trum diftindlly evolved, the folar image muft be extended to ten times its 

 breadth ; nor can it be imagined that rays fo attenuated could produce fuch palpable 

 efFe£ls. The Doctor, indeed, admits that a fmall thermometer was lefs affe£led. Yet it 

 is obvious that, if the aftion was due merely to the afflux of light, the fize of the bail 

 ought to occafion no alteration, fince, if it prefents lefs furface to the beam, it likewife, 

 in the fame ratio, expofes lefs furface to the furrounding air. A very curious reafon is 

 afl!igned — " the cooling agency of the afcending ftream of air, which would aft more 

 " ftrongly where the proportion of furface to bulk was greater*." The idea of an 



afcending 



• I muft iH fairnefs take notice that the words here quoted from Philof. Journal IV. jzi, are not Dr. 

 Herfchel's but mine, as tlie abridger or narrator of the contents of his paper. The Doftor's obfervation 

 (Philof. Tranf. Y. 1800. page i6i.) ftates in very general terms that the cooling caufes may, he fuppofes, 

 have a ftronger affeft on the fmall than the large thermometer. I do not here quote more fully, becaufe the 

 point feems not of fuch importance as to demand it, and becaule the only miftake refpefling it appears to 

 Ue with the ingenious author of the prefent communication. His fecond refleftions will not, furely, permit 



him 



