1825.] 3f. Berthier on Forge Scales. 133 



to suppose, since, when pure, they give no sensible quantity of 

 hydrogen by the action of acids. Besides, if they contained 

 half their weight of iron, they would give 0*85 of fused iron by 

 the assay, which is far beyond what we obtain. 



It remains to be ascertained if the forge-scales can be a 

 mixture of protoxide and deutoxide. If that were so, since the 

 protoxide is very greedy of oxygen, they should have a great 

 tendency to combine with that body ; whereas, not only are they 

 wholly unalterable by exposure to the air, but are acted on even 

 by concentrated and boiUng acid only very slowly and with 

 great difficulty. I endeavoured to determine their composition 

 by this method, estimating the quantity of oxygen absorbed 

 by the increase of weight ; but I was unable to convert them 

 entirely into peroxide. It is, moreover, very doubtful, if the 

 protoxide of iron can exist in a free state ; for being a base 

 which has such attraction for oxygen, that it decomposes water, 

 it is very difficult to obtain it absolutely uncombined. The dry 

 way appearing to be the only means by which we can hope to 

 succeed, I made several trials after that manner, but without 

 success. The following process seemed the most likely to ac- 

 complish the object in a direct manner. 



I took several black lead crucibles lined with charcoal, and 

 placed 100 grammes (1544 grains) of pulverised and finely 

 sifted forge-scales, in each; I then filled the crucibles with 

 charcoal, and closed their mouths with covers, carefully luted 

 on, and exposed them in a wind furnace to a heat of about 70° 

 of the pyrometric scale. I took them out of the fire in suc- 

 cession — the first in half an hour, and the last in three hours, 

 and compared the results. All the buttons had become solid, 

 without changing their form or diminishing in volume ; they 

 were covered with a coating of metallic iron, and the oxide in 

 the centre was neither fused nor altered; it gave the same 

 relative proportion of peroxide and protoxide by analysis \ik 

 humida, as at first. The thickness of the metallic coat was 

 proportionate to the time the crucible had remained in the fire ; 

 the maximum thickness was five millimetres (nearly '2 of an 

 inch.) It has a peculiar aspect; its surface is dull, and fracture 

 granular ; its colour is grey, inclining to olive ; it takes a 

 briUiant polish by being rubbed against hard substances ; it 

 may be cut with a knife, and reduced, in that manner, to a very 

 fine powder ; it is as soft as lead, and has no elasticity; it fiat- 

 tens by a blow, and retains the mark of the hammer ; its spe- 

 cific gravity, at the utmost, does not exceed one-third that of 

 ^tged iron; it is, in fact, pure iron, minutely divided, and in a 

 state analogous to that of spongy platina. 



If the cementation has been continued for a considerable 

 ^jHa4 #^! section of the button presents, from the surface to 

 ^§i&§i|^lmf 4r^t;,:.%,T§Ty,,^->yer of met^^^^^ of a deep 



