Experiments on two varieties of Iron. 101 



a length of 7.43 inches, while those on the inside had but 4.8 

 inches. The difference 2.63 inches is to be attributed to the 

 combined influence of compression and elongation of the interior 

 and exterior fibres. By measuring and marking bars before and 

 after bending them, such differences may under certain restric- 

 tions be employed as means of determining the positions of neutral 

 axes. Changes analogous to those already observed in the form 

 of the cross section, were also remarked in the present instance. 

 The next trial was by turning a right angle when hot on a short 

 portion of the bar, and subsequently folding another part over fiat 

 upon one of its faces. All the phenomena of developing curves 

 out of the square cross section were beautifully exemplified, and 

 the soundness of the iron when thus tested at a red heat, incon- 

 testably proved. A short portion of this inch-square bar was next 

 heated to a fair working; temperature and perforated with a punch 

 five eighths of an inch in diameter. No signs of cracking on the 

 sides or splitting longitudinally were observed. 



Four bars were then prepared, in all respects similar to the first 

 four taken from specimen No. 1, and respectively treated in the 

 same way preparatory to a trial of their tenacity. 

 The bar marked V, was completely annealed. 



" " VI, '' hammer-hardened. 



" " VII, " hammered till cooled to a dull red heat. 



" " VIII, " upset in the middle and annealed. 

 As the upsetting of No. VIII, had of necessity increased the 

 thickness of that part to which the operation was applied, care 

 was taken to reduce by filing, the cross section in the middle of 

 the upset portion, to less than that of the rest of the bar, in order 

 to insure a fracture in metal actually in that state. The trials on 

 No. IV, had led to the supposition of an increase of strength by 

 the process of upsetting, contrary to an opinion entertained by 

 some practical men. Experiment No. 2, on bar No. VIII, in the 

 following table, in which the fracture took place at the filed 

 section, gives not only the highest result on that bar, but also 

 higher than any other obtained from this variety of iron, except 

 those derived from the hammer-hardened bar, No. VI, and conse- 

 quently confirms our previous deduction. 



