ON THE CHEMICAI, NATURE OF CAST IRON. 13 



are so loosely held togcthei* that tiie various rates of cooling will determiuc 

 their combination or decomposition. 



From the analogy of tlic alloys of carbon and iron with other aUoys, it is 

 possible that the case is as follows ; namely, that the carbon and iron exist 

 in the molten state as a solution of the one in the other — when in the white 

 state as a sohdified solution of carbon iu iron, and when in the grey state as 

 a mechanical mixture of carbon and iron. 



If the above assumption be correct, then the conversion of white into grey 

 cast iron, and vice versa, may be easily explained ; for if in the liquid state the 

 carbon and iron be only a solution of carbon and iron, then we can easily 

 imderstand why the carbon, when the molten mass is heated to a high 

 temperature and cooled slowly, crystallizes out, and when cooled quickly, not 

 having time to crystallize out, remains homogeneously diffused through the 

 iron ; as it is very probable that molten iron will dissolve more carbon at 

 higher temijcratures than at lower ones. 



The two chief reasons for assuming that the carbon and iron exist in the 

 state of chemical combination are, 



(1) That when white iron is dissolved in dilute acids, the carbon combines 

 with the hydrogen, forming carburetted hj-drogen. 



(2) That different carbides of iron have been isolated. 



First, respecting the chemical behaviour of the carbon in the white 

 iron, the following remarks may be made. If it be assumed that the carbon 

 in cast iron is homogeneously diffused through the mass, then it must be in 

 an exceedingly fine state of division — in fact in just the state for combining 

 Avith other bodies; for it is well known that carbon as well as other sub- 

 stances possess properties, when in a state of fine division, which they do not 

 possess when in a more coherent form. Thus in the case of carbon, the more 

 porous it is the more active it becomes in dechlorizing liquids and absorbing 

 and condensing gases in its pores. Platinum is also a good example of this 

 fact ; in a dense form, as in foil, it only possesses the property of condensing 

 gases on its siirface to a very feeble extent ; in its spongy form it possesses 

 this property in a very marked degree, and as platinum-black possesses it to 

 a far greater extent than the spongy modification. 



As another example of the influence of the chemical activity of bodies when 

 in a fine state of division, we may take the eases of iron, lead &c., these 

 metals, when in a coherent form, undergoing when exposed to the atmosphere 

 only a very slow oxidation, but when in a very fine state of division (reduced 

 oxides) combining with oxygen instantly. 



The chemical behaviour of the constituents of alloys is sometimes diifei-ent 

 when in an alloy from what it is when alone ; thus platinum y^hon. alloyed 

 with silver dissolves to a certain extent in nitric acid. Now, taking the de- 

 terminations of the eonducting-powers of alloys as a means of testing their 

 cliemical constitution, I should with certainty say that there exists in the 

 alloys of platinum with silver no chemical combination, for their observed 

 and calculated eonducting-powers agree together better than for any other 

 series of alloys. The curves representing the eonducting-powers of the alloys 

 of these metals with one another possess the typical form of the alloys of that 

 class of metals to which these belong. 



2ndly. With regard to the definite chemical combinations which have been 

 isolated from cast u'on, it may be remarked that definite crystalline forms with 

 alloys do not necessarily indicate chemical combinations. Cooke*, I believe, 

 * Memoirs of the American Academy (new series), vol. viii. p. 27. 



