98 2\'eiif Vrcccfs for mahing 



round before the point had fet up all round like lead. The 

 fame piece was afterwards fubjefted to feveral trials of harden- 

 ing, bendino-, and filing, in all which it manifeftcd properties 

 peculiar to the I'oftell n)alk'able iron. ^The dcnfity conveyed 

 tn the nu'lal bv fufion, feems to form an univtrfal charafter- 

 illic feature of this fpecics of metal. When the lile is applied 

 to the melted mafs, it meets with a much greater degree of 

 rcliltanct' than from common malleal)le iron j but this exifts 

 no loncer after being forged, and the file then acts with the 

 greatcd poHible efietT. Grains. 



Kxp. \'III. Pieces of Swedifh iron weighing 1093 



Itnlian marble' - 2180 grains, 



Stourbridoe clay, old pot 2180 gra-ins. 

 This proportion is lix to one of that prefcribed bv 

 C. Clouctj and from which gray caft iron (hould 

 have been obtained. Ip 45 minutes I deemed the 

 fulion complete, and withdrew the crucible. When 

 cold, I found a very perfecl: metallic button relling 

 beneath a great thicknefs of glafs. The external ap- 

 pearance, in every point, refcmbled that of No. i. 

 It weighed _ _ _ 1070 



Loft in fufion ' 20 

 et]n3l to- — - part the original weight of iron. The glafs in 



this experiment was feveral Ihades lighter than that obtained 

 in No. 7. The realon for this feems obvious, as feven times 

 the quantity of glafs pofll-lfed only a quantity of colouring 

 principle fimilar lo the former, if bv this be underflood the 

 fpecilic quantity of iron loft in fufion. The fra^lure of 

 this button was preeifely that of No. 7. One-half of it was 

 drawn into a finall cutting chilel, and hardened at a pretty 

 bright red heat. It ftialed a little, and felt, for a few ftrokcs, 

 perceptibly harder under the file. Upon applying the ehifel to 

 cut a piece of newly hammered Swcdifli iron, it made a eonfi- 

 tferable inipreflion with a moderate blow from a 3 lb. ham- 

 mer; but the whole edge was flattened, and at a fecond blow 

 the edge Itavcd completely, and nieafured nearly i-8th part 

 of an nub acrols. 



A fragment of the iiime experiment was fub)e<5led to a 

 variety of lefts, in none of which the nioft diftant fymptom 

 of (ieel was manifeft : on the contrary, in point of ftrenglh 

 and uur.il)ilily it was found inferior to pieces of the original 



bar. Grnns 



Kxp. IX. Swedifh iron, in pieces - ^099 



Italian marble - 1099 grains, 



Thi=i 



