« PreviousContinue »
As an example of the curve of temperatures obtained, see fig. 3. At the abscissa marked 0 min. cold water was turned on, and at 15 min. steam. The points marked are experimental points. There are no points marked at 0 and 15 min. as I had to be in another part of the room at those instants. It will be noticed that the cold water or the steam (as the case may be) has done nearly all it can do, before the other is turned on, especially in the case of A. This can also be seen from the numbers given below.
The following are the details for each rod:—
Rod I. (inch).-Diam. = 3377 cm. Density=849. Specific electrical resistance 7.24 x 10-6 ohms. 7 cms. between junctions. Time-period 10 min. Determination of temperature every half-minute.
The results, in C.G.S. units, calculated from these numbers
* At the end of this set the joint B was found displaced through about 2 mm. away from A. This introduces a possible error of about 23 p. c. in the distance, or 5 p. c. in the conductivity, while leaving the emissivity practically unaffected.
Rod II. (inch).-Diam.=6426 cm. Density 8.47. Specific electrical resistance = 7.56 x 10-6 ohms. 10 cms. between junctions. Time-period 20 min. Determination of temperature every minute.
Rod III. (inch).-Diam.=9654 cm. Density = 8.49. Specific electrical resistance = 7·71 x 10-8 ohms. 12 cms. between junctions. Time-period 30 min. Determination of temperature every minute.
The third sine term is, in the case of B, too small to give reliable results, but they have been calculated in some cases, and curiously enough the conductivity always comes out lower than that given by the first term. The following have been calculated :
Taking the mean of the chief results obtained, we get the following table :—
It hence appears that the emissivity of a body depends to a large extent on the form and dimensions of that body, and that the formula for conduction of heat along a rod, as usually given in text-books, can only be used in a very restricted
It will be noticed that the two emissivities (deduced from the "a" term and the sine term) are to one another approximately as 1 to 12. The ratio appears to be too uniform to be due to accidental error of experiment, but the meaning of the two values is not quite clear. In order to obtain further information, the inch rod was heated at one end by steam under approximately the same conditions as before, and the E.M.F. of a german-silver-iron couple at different points on the rod determined by the compensation method in the ordinary way. A preliminary set was taken, which gave about the same results as the following more accurate set (readings were taken going down the rod and then going up):
Phil. Mag. S. 5. Vol. 39. No. 238. March 1895.
Temperature of air about 100.
1 cm. of bridge-wire corresponds to rather less than one degree.
Dividing the above into two lots, and taking k=2386, get (this method is sufficiently accurate for this purpose) :
The mean temperature of the rod in the experiments by Ångström's method was about 27°, and the two emissivities found were 000207 and 000242, so that the steady-state emissivity comes just about halfway between the two. The rod, however, had become somewhat tarnished during the summer, and had to be cleaned again, which may or may not have made some difference. Further experiments will be needed to settle the point conclusively.
When the last part of the experiments was in progress, Mr. A. W. Porter, B.Sc., drew my attention to Messrs. Ayrton and Kilgour's paper before the Royal Society on the variation of emissivity with the diameter of the wires experimented upon. The results obtained corroborate one another, and also corroborate results that had been obtained some time ago by Péclet, and which are recorded in his Traité de la Chaleur. Péclet appears to have carried out a large number of experiments on emissivity, the best account of which is found in the 3rd edition (not the 4th) in vol. i. and a long note at the end of vol. iii. (published in 1860-61).
In conclusion I must express my obligations to Prof. Foster and Mr. Porter for much help given in the course of these experiments.
University College, London,
Dec. 29, 1894.