different Earths for Carlo?!. 35 



Were we to proceed to reason on what this variety of 

 result depends, and referring to the experiments formerly 

 commimicated, we should have expected that clay and 

 silcx, each absorbing a considerable portion of carbon, 

 would have required a greater dose in the experunents with 

 oxide to have let tall the first portion of iron ; and that, as 

 calcareous earth betokened no affinity manifested in a si- 

 milar manner to carbon, ores thus compounded would have 

 let fall their metallic contents with the most minute com- 

 parative quantity ot carbon. 



The reverse of all this turns out to be the fact : for the 

 aro-illaceous and siliceous compounds separate iron with 

 the smallest portion of carbon ; the calcareous compound 

 requiring three times as much. . . 



Hitherto we have discovered no direct active principle m 

 calcareous earth, acting as a stimulant to the existing affi- 

 nities betwixt carbon and iron, beyond facilitating, almost 

 mider every circumstance, the perfect fusion of the com- 

 pound. If we suppose it to remain neutral in this respect, 

 then, to explaiii the phaenomenonof the argillaceous and si- 

 liceous compounds, we must suppose an active principle 

 exerted by each of these earths, nearly in the same ratio, 

 upon the oxide of iron, decompounding the oxide, and 

 either liberating the oxvgen or uniting with it. The re- 

 moval of this immediately constitutes an affinity betwixt 

 the particles of iron and the entire portion of carbon, the 

 donsequence of which is the revival of the iron. 



That these conjectures are well founded, may be gathered 

 from experiments similar to the following : 



Calcareous earth six parts, or - - 120 grs. 

 Oxide four parts, or - - - 80 



Charcoal I- loth of the whole - - 13]. 

 To this were added of Cornwall clay - 120 

 From the fusion of this mixture a neat metallic spherule 

 was obtained which weighed 10 grains, and was equal to 

 5 per cent. This is the same, with regard to proportion, as 

 Experiment IV, third class, with the addition of 120 

 grains clay; and it appears a fair consequence that 1.^. per 

 cent, of more iron was revived. 



There appears still another way by which we may in part 

 account for the eailv separation of" iron from the argilla- 

 ceous and siliceous compounds. 



In a former part of this inquiry it was shown that the combi- 

 nation ofcarbun with clav and li'ilexdidnct produce full cfToct 

 unless the latter entered into fusion ; and it is evident ironi 

 the foregoing experiment^*, that the compounds formed with 

 ^ ^ C 2 o.^tide 



