Cnft Steel from Bar Iron. lOI 



Their fimilarity and identity with thofe dcfcribcd In Nos. 7, 

 8, and 9, rendered fiich a defcription unnecefiary. Suffice 

 it to fay, thev were applied in the fame (liapes to the fame 

 fpccics of labour, and found equally inapplicable to the pur- 

 pofes for which they were intended. 



Neither iij the experiments recorded in this paper, nor in 

 thofe which accompanied the former communications, was 

 there found anv product refeml)ling fteel, or poflefled of any 

 of thofe properties which iron derives from its combination 

 with carbon : on the contrary, the refulting produft in all of 

 them was found debafcd, in point of ftrciiglh and elafticity, 

 below the ftandard of common malleable iron. The uniform 

 quality of the various metallic mafl'es, though fufed in contaft 

 with widely dilferent proportions of earths, leaves little room 

 to doubt that the latter are entirely neutral ; nor docs it ap- 

 pear, upon a review of the experiments performed with a 

 mixture of carbonate and argil, that the lois of metal is de- 

 pendent upon the quantity of earths fufed in conlaA with it, 

 as the following table will Hiow : the fame review exhibits a 

 triple lofs of metal in thofe experiments wherein the car- 

 bonate of lime was fufed alone with the iron. 

 Experiment No. i. lofs was one 73d part. 

 2. - - 70th 



Avcraoe lofs in fufion — one 56th part. 

 The three firlt fufions feem to have experienced a fmallcr 

 lofs in metal than the fubfequent ones. This I am apt to 

 attribute to a comparatively imperfe(5l fufion, as the buttons 

 obtained in Nos. i, 2, and 3, were conliderably pitted, and 

 every way lefs denfe than thofe afterwards produced, when f 

 had learnt tr) regulate the necellary heat to a ccrtaintyi 

 Jlxjierimcut No. 6. lofs one 17th part 

 9. - 27th 

 12. - 17th 



Average lofs in fufion one 20 |d part. 

 It appears alio that tlie quantity of carbonate ufed, and the 

 lofs of iron fuOained, were not analogous to each other, 'i'o 

 vcrifv this, No. 6 and 12 may be compared, 



CJj In 



