18 The Rev. T. R. Robinson on the Relation between the Temperature of 



containing 5.5 cubic inches of water, a current = 3.527 was passed for twelve 

 minutes, the air and water being both nearly 72°. The thermometer rose 77°.5, 

 and the resistance A = 257.6. The wire was almost white, and its temperature 

 must have been near 1500°. The instrument was then cooled, filled with water 

 (which, by incUning it, was made to fill B), a resistance = 165.6 added in the 

 circuit, and a current = 3.558 passed for twelve minutes. Now the thermome- 

 ter rose only 29°. 7, and the resistance = 89^0. Here all was the same, except 

 the increase of A by the heat ; and that determined the greater heating power 

 in the first experiment. The second was repeated, but without any inter- 

 posed resistance in the circuit: bubbles of steam formed round the wire, 

 which were condensed with sharp snapping ; the current = 6.045, which would 

 have melted it in air; the resistance = 94.5 ; and the rise of the thermometer 

 = 83°.2. 



From this it follows that/(t/) = a (1 -H ry). 



The function f'{y) is far from being so easily determined. Fourier and 

 Poisson, in their celebrated investigations, have assumed it = hj, following New- 

 ton. This, however, is quite at variance with observation. Dulong and Petit, in 

 their memoir, have assigned expressions for the efiect of the air and radiation, 

 which represent their observations very exactly. According to them the effect of 

 radiation is as 1.0077^-1, andthatof theair as7''•^^ The heated body which they 

 employed was the bulb of a thermometer enclosed in a globe of copper a foot 

 in diameter, blackened on its interior surface, and kept at an invariable tempe- 

 rature. The highest range was under 500°, and of course far below the point 

 at which light is given off, — an agent which, doubtless, interferes with heat. 

 Accordingly, their law of radiation fails altogether in my experiments, but the 

 other is nearly exact. We are enabled to infer this from the law already men- 

 tioned, that the heating power of the current is as its square multiplied by the 

 resistance.* Now in the experiments Nos. 1-14, the wire is in contact with 

 air, while in Nos. 15-22, which were made in vacuo, the effect of that medium 



• This law has been often verified, but the experiments just described gave a good illustration 

 of it. In the first and third, the ratio of the heats excited is 1 : 1.074; while that of .4 C^ is 

 1 : 1.077; the difference arising from this, that, the temperature being higher in the last, more 

 was lost. It may also be mentioned as evidence of the heating power of that current, that it gene- 

 rated in the wire as much heat as would have ignited eight ounces of platinum to strong redness. 



