92 report — 1877. 



vapour (at a tension of a few inches of mercury, and at a temperature of 

 110° F. to 130° F.) is as efficacious as water itself, no special experiment 

 with the thin wire thermopile to solve the question as regards dry air alone 

 was made in further trial of what would he, if found to he successful, a 

 practically very valuable simplification. 



The other variations introduced in this year's series of experiments were 

 to increase the temperature differences and the heat-flow through the tested 

 plates by removing all but the most necessary sheets of caoutchouc lining 

 between them and the heater or cooler, and by raising a more rapid supply 

 of steam in the heater with a stronger flame. The actual temperature of the 

 plates was also raised in 6ome experiments by shifting all the movable 

 linings from underneath to above the plates so as to bring the latter further 

 from the cooler and nearer to the source of heat. The success of these ex- 

 periments was only partial, because, in the strong temperaturo differences 

 which prevailed (when temperatures between 130° F. and 160° F. were 

 noted in the water-bath), false currents arising from want of homogeneity 

 in the heated wires presented themselves in the thermopile, which, in spite 

 of the number of its coils, did not neutralize each other entirely ; and it was 

 found necessary to test it very carefully (as described in the last Report) 

 by introducing a hot paper-covered thick plate of iron between the lappets 

 of the thermopile in place of a rock-specimen, and observing the temperatures 

 of its two faces at the water-bath. Errors of indication of the thermopile 

 were thus discovered, and were noted at different temperatures of the iron 

 plate, arising from the abrupt changes of temperature along the wires. The 

 way in which the temperature of the water in the bath was changed, quickly 

 or slowly, from hot to cold, or vice versa, seemed especially to influence these 

 considerably ; and it was finally resolved to abandon the attempt to obtain 

 new results, by these means, of the thermal conductivities of the rock-plates 

 at higher temperatures and under very different circumstances of the heat- 

 flow through them from those which had been employed before, although the 

 known allowances for the small erratic deportment of the thermopile always 

 gave under these entirely new conditions results which did not differ appre- 

 ciably from those which were previously observed, and which have been re- 

 corded in the earlier tables of these Reports. Temperatures of the rock- 

 faces between 100° F. and 120° F. were found by trials with the iron plate 

 to be easy to observe correctly in the water-bath, with proper care in its 

 management, with errors not exceeding two or three tenths of a degree, 

 while the temperature differences requiring to be thus observed varied from 

 between 3° and 4° with quartz to between 30° and 40° with shale and sand. 

 About this range of temperature of the rock-faces was accordingly adopted, 

 by properly thickening the lining between them and the heater, in the expe- 

 riments which afforded the following table of results (p. 94). While it would 

 be necessary, in order to deliver the individual wires of the thermopile from 

 strong effects of temperature-differences, and to obtain scientifically accurate 

 results, to discard steam-heating and the use of temperatures much above 

 those of the outer air altogether, resorting for, example, rather to cold water 

 from a main to produce the temperature difference necessary to transmit heat 

 through the rocks, or using water otherwise cooled artificially in the cooler, 

 and exposing the under surface of the rock with a lining and a metal plate 

 to the ordinary temperature of a room, yet with the small uncertainties 

 which, without doubt, remain in the indications of the thermopile from the 

 cause here pointed out, in this and in all the earlier tables of absolute con- 

 ductivities and resistances which the Committee has appended to its Reports. 



