TRANSACTIONS OF THE SECTIONS. 49 



series as before, the following numbers on which the comparison may be made are 

 obtained : — 



Assuming, for the purpose of comparison, that the highest evaporative power in each 

 of the four series was truly represented by the lead-reducing power of the same coals, 

 and supposing that the evaporative powers of the six lowest of the series were to be cal- 

 culated from their reductive powers, we should obtain the following proportions : — 



1. The American series 30-03 : 10-18 : : 25-76 : 873 



(=: calculated steam power) 7-55 



2. The first British ditto 31-0 : 10-59 :: 26-7 ; 9-12 



(= calculated steam power) 9-72 



3. Second ditto, 33-7 : 10-51 : : 26-2 : 8-88 



(= calculated steam power) 7-14 



4. Third ditto, 32-1 : 9-90 : : 27-1 : 8-24 



(= calculated steam power) 7-09 



Hence it appears that in attempting to compute the evaporative power of coals which 

 s;ave the lowest results from their lead-reducing power, as compared with that of coals 

 giving the highest evaporative action, we obtain in every one of the four cases num- 

 bers largely exceeding the practical results from the steam-boiler : — 



Calcn- ISxperi- j.j~ 

 lated. mental. 



The first comparison gives... 8-73 - 7-55 = 1-13 = 15-3 "l g„t excess above 



The second comparison gives 9-12 — 7-72 =: 1-40 = 18-2 I ^,, " *. , •■ 



rr,, ,1 ■ J ^. ■ o oQ 17 T A 1-7 1 o/f o T thc cxperuiieutal rc" 



The thu-d comparison gives 8-88 — 7-14 =: 1-74 = 24-3 , *^ 



The fourth comparison gives 8-24 — 7-09 = 1-15 = 16-oJ 



From this it should seem that in applying Berthier's test as a standard of compa- 

 rison, we are liable to over-estimate the evaporative power of the coals of high bitu- 

 minousness (containing much hydrogen), and that the excess may, in extreme cases, 

 amount to from 16 to 24 per cent, of their pi-actical efficiency. It is not, however, 

 to be inferred that the series of analyses by this method is without practical value. 

 Having so wide a range in the constitution of coals with corresponding reductive and 

 evaporative powers already determined, we may readily intercalate any new variety 

 of coal analysed and tested by litharge, and thence deduce its approximate evapora- 

 tive powei-, subject, of course, only to that degree of uncertainty which results from 

 a want of perfect conformity in constituent elements with the coals placed nearest to 

 it in the order of reductive powers. When both proximate and ultimate analyses 

 show a good degree of coincidence with any coal already tested under the steam- 

 boiler, and the lead-reducing power of the new variety also corresponds nearly with 

 the mean of those above and below it in the scale, there seems to be little doubt that 

 its evaporative power will also form a corresponding term in the series of calorific 

 efiiciencies. 



It is generally known that several chemists have proposed the Jixed carbon of coals, 

 or that remaining in their cokes, as a standard by which to estimate their evaporative 

 powers. This method is subject to the objection that the weight of coke itself, and 

 consequently the weight of its fixed carbon, is liable to vary within certain limits, with 

 the rate of distillation, that is, with the rapidity of the coking process. In many 

 highly bituminous coals the sudden application of an intense heat will expel, in the 



1851. 4 



