1^8 EstpentntnU on the ConduSling Poivers of Fluids mth Ragiird to Heat. 



But as the caffe is not fo fimple, but muft neceffarily include the veflel, or folid coating of 

 the fluid mafs, the confequences will be different. If the fluid be capable of receiving and 

 giving heat only from and to folids, vft mud contemplate the heated parts as rarefied, (that 

 is to fay increafed in bulk) and confequently diffuGng themfelves laterally over the reft of 

 the furface 5 as coming into conta£t with and heating the veflel ; as urged to defcend at the 

 circumference by the fucceflive arrival of other more heated particles, and confequently 

 converging again beneath, and afcending near the center : we muft confider the folid fide 

 as conducing the heat ftill farther down ; and there caufing an afcending current to joi% 

 its water to the converging ftreams above. So that the water may be confidered as receiV'^ 

 ing heat from a folid in contact with its furface near the center ; giving it again to a folid 

 •t its circumference ; and this laft folid in its turn heating another ftratum of water be- 

 neath. As the veflel fliould become heated to a greater diftance downwards, the effeft 

 would be more perceptible to a greater depth in the fluid ; and from the want of commu- 

 nication between the particles among each other, the afcents and convergences might have 

 different velocities, according to the temperatures at the feveral depths, and preferve dif- 

 ferent temperatures in the feveral ftrata regularly increafing according to the advancement 

 of the operation. 



We thought it probable, as the pafl'age of heat downwards is very flow, that thefe cur- 

 tents, if they exift, might alfo be fo flow that difi^erent temperatures might be found at one 

 and the fame time in the fame ftratum ; and w,e fuppofed the experiment would be more 

 Conclufive, if the heating folid were made to cover the whole furface of the fluid, and to 

 reft upon the vefl"el itfelf. For in this arrangement, as the upper edge of the veflel would 

 be as hot as the furface of the fluid, we concluded that there could be no divergence, but 

 that a fimple afcending current would be produced near the, fides, with a defcending one 

 near the axis of the veflel. We chofe a veflel of wood, as the worft condudlor we could 

 make, * for obvious reafons. 



Plate X reprefcnts the apparatus, A B is a ftage for philofophical experiments contrived 

 by Profeflbr S'Gravefande, C is a cylindrical veflTel of mahogany of the following dimen- 

 fions; outfide diameter 4.75 inches, outfide height 2.75inches, inflde diameter 3.5 inches, 

 infide depth 2.0 inches. In the upper edg?, which was confequently 0.625 inches thick, 

 there was a cylindrical excavation to the depth of o. i inches, and of the diameter of 4.5 

 Inches. In the bottom of this veflel were fixed with fealing wax, two thermometers L and 

 M, the one in the axis, and the other about 0.2 inches from the infide furface, and the 

 bulb of each ftood 0.75 inches higher than the inner furface, or bottom of the veflel ; con- 

 fequently the clear depth of the veflTel from the upper edge to the thermometer, was 1.25 



* A cylififlerbf glafs 1 inch in diameter, and 0.25 inches thick, was placed on the horizontal cover of 

 8 veffel containing-water kept boiling by alamp, and a like cylinder of light dry mahogany was placed befide 

 it at the fame inftant. A very minute cone of tallow (tood on the center of each piece. That upon the 

 glafs f«Jl down by fufiop in one minute ; that upon the wood in two. This method is much more apcurat* 

 Xluin that of Franklin with long rods defcribed in Dr. Ingenhousz's Effays. 



inclieSi 



