.^6 Jf'iUiitg tf Steel. — Fltxun jf Metallk Bars. 



lover (tegree of heat in which it may be foft enoagk to naite \t'ith iron, yet without ha« 

 zard of running under the hammer ? A few expcviments decided tlie queftlon •, for the 

 facl is, that caft-fteel in a white heat, and iron in a welding Iieat, unite completely. 



It mud not be denied that confidcrable nicety is required in giving a proper heat to the 

 ftecl ; for, en applying it to tlie iron, it receives an increafc of heat, and will fomctimcs 

 run on that increafe, though it would have borne the hammer in that ilate in which it 

 was taken from the fire. 



I need fcarccly obfervf, that when this procefs is intended, the (leel and iron mud be 

 heated feparatcly, and the union of the parts propofed to be joined, effecled at a finglc 

 heat. In cafe of a confiderable length of work being required, a fuitable thickncfs muft 

 be united, and afterwards drawn out, as is pra£lifed in forging reap-hooks, &c. 



The fteels on which my experiments liave been made, are Walker's of Rotherham, and 

 Huntfman's, between which I difcover no difference -, and though there may be fome 

 trifling variation in the flux ufed for melting, they are probably the fame in efl'entials. 



2. Flexure of Compound Metallic Bars ly Change of Temperature. 



A BAR of fteel upwards of fix inches long, 0.56 inch broad, and 0.05 inch thick, was 

 hard-foldered * to a barof brafs of the fame dimenfions, but twice the thicknefs. This 

 compound bar was fixed at one end to a finiple metallic bar, as in fig. 2. Pl.V. (facing p. 96) 

 and at the other end there was a Hiding piece to meafure the deviations by flexure. It 

 muft be obfcrved, that a thin piece of metal was placed between the fimple and compound 

 bars, at the pinned part A ; fo that the bars were not permitted to come into contact in 

 any temperature, but each pofition was determined by the Hiding piece, by means of a mi- 

 crometer- fere w and magnifier applied to it whenever its place was required to be known. 

 The ck'ar part of the compound bar from the pin A, to the end D, was exactly fix Inches. 



By heating the apparatus from 66° to 212°, namely 146", the catch was moved through 

 129 parts of the micrometer, each of which was TiVs'^ "f an inch. The quantity 129 was 

 therefore equal to 0.0921 inch. 



Both fides of the compound bar were then carefully filed away till the metals were each 

 (as near as could be afcertained) only half their former thicknefs. The flexure, by a like 

 difference of temperature, as in the former trial, now proved 278 parts, or 0.1 99 inch. 



The bar was then filed to about .|-ths of its original thicknefs, ftill preferving tlie rela- 

 tive thickntffes of the metals. Its flexure by the fame heat as before was 313 parts, or 

 0.224 inch ; in which, however, from the fpring of the thin bar, it was doubted whether the 

 catch had been moved to the fjme diftance as the bar itfelf might have moved alone. 



Onconfidering thefe deviations, the numbers prove to be inverfely as the thickneffes of 

 the bars, with no greater error than might arife from the meafures by which the real 

 thickneffes were afcertained. Thefc experiments may therefore be faid to agree with the 

 theory f. 



' For the information of thofc who may not be acquainted %vi»h the proceffes of foldering, I muft ubferre, 

 that this operation confifts in uniting the pieces by tlic fufion of a kind of brafs, which contains an citraordi- 

 n«ry proportion of linc, and flows into the joint before the heat is fuSicient to ttwit the pieces. This kind of 

 work will bear hammering and bending, becaufe the folder is not brittle. Soft-foldering is erteftcd by plumbert* 

 folder or tin, which forming a kind of bell metal with the copper is brittle in the joint, tind thcrcfcrc unfi^for 

 work intended to fuftain either blows or flexure. 



+ Philofophical Journal, I. 6*. 



