TRANSACTIONS OF THE SECTIONS. 225 



Fahrenheit degrees by inserting a double tin lid between the corks, under the two 

 opposite faces of which water of different degrees of temperature was made to cir- 

 culate, and the temperature of the water was made Ivuown by thermometers inserted 

 in the lids. The other arrangement consisted in iixing the corks to the ends of a 

 pair of wooden tongs, so tliat the rock plate could be pressed between them as soon 

 as it was taken oft" the heater. It was in a first trial of this last arrangement that 

 no perceptible signs of heat-difference could be observed between the rock-faces. 

 To increase the actual difference, however, the edge of one of the stone plates was 

 surrounded with a band of paper, and the upper surface was then covered with 

 mercury, upon which the thermometer-flask was placed, this liaving also been filled 

 with m'ercurj^ instead of water to accelerate conduction. On taking the rock (a 

 plate of white marble) out of the apparatus after this treatment, and testing its 

 thermal difl'erence with the galvanometer, it was found that one surface was about 

 7° F. hotter than the other, while the flask containing 9 lbs. of mercury was 

 heated 1° F. in about ten seconds. This corresponds to the passage of 330 heat- 

 imits per hour through a 1-inch plate of the same rock (1 square foot in surface-area), 

 with the same difference of temperature on its opposite sides of about 7° F. For a 

 difference of 1° the transmission of heat in the same time would be 47 heat-units, 

 while the value obtained by Peclet for tine-graiued white marble was 28 heat-units 

 per hour. It is evident that some of the diflFerence of temperature between the sur- 

 faces of the plate subsided and disappeared in lifting it out of the heating-apparatus 

 and transferring it to the galvanometer, so as to make the conducting-power of the 

 plate appear to be about half as great again as its known value. The galvanometer, 

 which at first marked 7°, rapidly sank to zero as the rock was moved about between 

 the cork projections. 



The other disposition of the iron-platinum couples (on corks fixed to the heating 

 and absorbing plates) touching the rock-surfaces dining the heating operation, was 

 found to introduce errors in the opposite direction by showing, apparently irom the 

 conducting-power of the cork supports, greater temperature differences of the sur- 

 faces than can reasonably be supposed to have existed. Thus with the same plate 

 of white marble a temperature difference of 50° F. was recorded, instead of 7° F. 

 as in the former case ; while 264 heat-units per hour was the rate of conduction 

 through a plate of standard size for that difference, corresponding to only 5^ heat- 

 units for a difference of one degree, and not exceeding a fifth part of the value found 

 by Peclet. The same process was tried with the tAvo kinds of shale, and showed, as 

 before, that their conductiug-power is much less than that of fine-grained marble, 

 the quantities found for their conducting-powers being 2| and 2 heat-units per 

 hour, or less than half as great as that of marble. The heat-conducting power of 

 ordinary calcareous stone is similarly found by Peclet to be about half as great as 

 that of fine-grained marble, the hatter varying between 22 and 28, and the former 

 between 11 and 13 ; and the results of further trials will, without doubt, confirm 

 more closely the exact values which he assigns. 



Had time allowed the experiments to be repeated with a new arrangement of 

 the apparatus, the sources of error peculiar to each of the above methods would 

 have been readily removed, as their origin is in each case easily explained ; and 

 another series will be undertaken with the excellent collection of rock sections that 

 have now been provided for them. In drawing up this description of the first trials 

 to which they were subjected, it is sufficiently interesting to observe that not only 

 the relative values but also the absolute quantities of the heat-conducting powers 

 of different substances obtained by Peclet are approximately confirmed, since certain 

 kinds of stone are found to have less than half the conducting-powers of other kinds ; 

 and in the case of marble the quantity of heat passing through a square-foot plate 

 one inch thick per hour, with a difference of 1° F. between the opposite faces, was 

 found in two trials (giving the conductivity respectively in excess and defect) to le 

 between 42 or 47 and 5 or 7 heat-units, while the value of certain marbles found 

 by Peclet varied from 22 to 28 heat-units. The corresponding numbers obtained 

 by Peclet for certain metals, as copper, iron, and lead, are 515, 233, 113 heat-units 

 per hour, or many times greater than those of terrestrial rocks. The latter 

 occupy an intermediate place between the metals and such substances as the Aariou.s 



