Free Convection of Heat in Liquids

Free convection of heat from a hot wire in various liquids.—(1) Photographic study of the convection streams. Both a parallel ray and a schlieren method were used and gave similar results. Photographs are reproduced for water, olive oil and glycerine. In all cases there is a necking down of the stream a few millimeters above the wire. When the wire is first heated, a layer of liquid all around the wire is heated; the lower half first moves down, then stops and goes up joining the stream which follows the upper half, which had proceeded as a distinct semi-circular pulse. The appearance of the image of the wire tends to confirm Langmuir's suggestion of a stationary film on the surface. Bubbles on the wire in olive oil were observed to execute interesting oscillations. (2) Variation of heat loss with the excess temperature of the wire over that of the liquid. After the wire (a silver one,.33 mm in diameter) had been heated electrically for about 10 minutes a steady state was reached; then the resistance of the wire was measured and the mean temperature of the liquid was determined with a mercury thermometer. It was found in the case of water, alcohol, C${\mathrm{Cl}}_4$, glycerine and castor oil, that the heat loss per ${\mathrm{cm}}^2$ per sec. is equal to $b{\theta }^n$ when $n$ has the values 1.15, 1.105, 1.125, 1.25 and 1.24 respectively. The measurements of A. H. Davis are shown to give values of $n$ of about 1.15 for five liquids, but there is an indication that $n$ depends upon the diameter of the wire. The factor $b$ varies with the liquid and the mean temperature and also perhaps with the diameter of wire.