COAL AND IRON OF CUTTACK. 19 



This sufficiently shows the main constituents and proves the general 

 composition to be as I have stated. 



Some of the slags resulting from the rude operations of the native 

 workmen are not however so heavily charged with iron as this ; but 

 few of them contain less than 40 per cent, of iron. 



Assuming this, then, as the lowest or minimum percentage, and seeing 

 that the iron in the ores reduced exists in the state of peroxide, it is 

 obvious that, unless the ore used contain more than 36 per cent, of iron, 

 no iron whatever can be obtained by the process adopted, the whole 

 of the iron going into the slag. * 



The important question then arises, how are these processes to be 



Improyements necea- improved ? I have mentioned that the furnaces 

 **'^' now in use are all made of clay, and I have 



described the rude system of blast which is employed. It is, I think, 

 certain that neither could such means produce, nor such materials resist 

 the effects of, a greatly increased temperature ; and yet, if ordinary 

 fluxes be used, to increase the per-centage of produce, such increased 

 temperature is absolutely necessary. 



The tribasic silicate which now forms the slag is one of the most 



_. , . ,^g easily fusible of the siUcates (excepting those of 



necessary. q^q alkalies), while the sihcate of lime and alu- 



mina, the slag resulting from the English and European smeltings, 

 requires a very high temperature to render it fluid, and the pure silicate 

 of lime still higher (of course alumina could readily be added if needful.) 



A total change, therefore, of the system adopted must be introduced 



_, „. . before any great improvement in this direc- 



Cnange of furnaces, &c. jo r 



therefore required. tig^ ^^^^ result. 



* Kemembering that the average of the clay iron-stones, from which nearly all the iron 

 is produced in England, is not higher than 33. to 34 per cent., some idea may be formed of 

 the imperfection and extrayagance of the native process, . . 



