60 REPORT — 1879. 



which have the extreme specific heats by weight, 029 and 037. The 

 present well-measured specific heat of pnmice stone (024, unless the plate 

 contained a considerable quantity of hygroscopic moisture) , is also appre- 

 ciably above the common value to which the heat-capacities of nearly all 

 the different descriptions of rocks tested approximate very closely in the 

 Table. 



To the above-proved rule of partial water-absorption by boiling, among 

 the very porous rocks, the plate of pumice stone presented an exception. 

 While absorbing a fifth of a pound (75 - 6 per cent, of its weight) of water 

 by boiling and immediate immersion in cold water, which far surpasses 

 the observed porosity of any other porous kind of rock examined, only 

 three-quarters of an ounce, or 021 of the former quantity, enters the 

 plate, and occupies its pores during the process of boiling only. 1 The 

 fraction of half the total water gain, provisionally assumed in the Table to 

 be introduced into the porous rocks by boiling, is therefore here too great, 

 instead of too little, by about a half of its amount. The large uncertainty, 

 until this plate's water-absorption could be re-observed, led to the omis- 

 sion, in the Table of last year's Report, of the data found for pumice stone, 

 the real values of which are now given in the subjoined list of verified 

 determinations. 



In the hope of discovering an explanation of the wide difference which 

 exists between the various conductivities hitherto recorded in these 

 Reports, and a list of similar conductivities published in the Proceedings 

 of the Royal Society of Edinburgh in 1873 (vol. viii., p. 66), by Professor 

 G. Forbes (the values in which are not more than a fourth or a fifth of those 

 described in these Reports), the Committee requested Professor Forbes to 

 search for possible errors among the numbers used as constant factors in 

 his calculations, while it submitted its own reductions to a similar ex- 

 amination. The result of Professor Forbes's re-examination is not yet 

 received ; but the Committee has had the annoyance to find that one such 

 small error has unsuspectedly been committed in its own determinations. 



Among the factors used, since the outset of its experiments, to convert 



into terms of absolute conductivity the rate of heat- flow measured directly 



in the 5-inch plates, a number, 196, was used inadvertently in place of 



the correct multiplier, 220, to effect a portion of the transformation. All 



the observed values that have hitherto been described in these Reports as 



obtained from year to year of the absolute conductivities (k), and of the 



• & 

 ratio - of the various rock-specimens which have been tested are there- 



1 Equally irregular departures from perfect conformity to a common rule occur 

 in some examples of the less porous kinds of rocks, where the yery moderate absorp- 

 tion of water, however, renders the deviations of their properties of little sensible 

 influence, as affecting the provisionally assigned values of their specific heats so as 

 to make them needful of any appreciable corrections. The hot-water absorption by 

 gas coke is like that of pumice stone, but a quarter, instead of a half or three- 

 quarters of its total water gain, which in this slightly porous substance is only 2-9 

 per cent, by weight. The small correction which this entails on the specific heat 

 by weight (0193), as given in the Table, is the additional quantity 00073, making 

 the real specific heat from the experiments, 02003. This is even more nearly iden- 

 tical with the value, - 201, for coke of anthracite, given by Eegnault, than the 

 former provisional value was, the near agreement of which with Kegnault's deter- 

 mination was pointed out in the comparative Table of such observations in last 

 year's Report. 



