ON THE TREATMENT OF STEEL FOR THE CONSTRUCTION OF ORDNANCE. 399 



indeed. Two welded coils of equal dimensions were made from this 

 material, and there was no appearance in either case of defect in the 

 welding. Both of these coils were tempered in oil, and one of them was 

 applied as a jacket to a steel cylinder closed at both ends and used for the 

 purpose of determining the pressure exerted by different charges of gun- 

 powder fired in confinement. An exact duplicate of this cylinder was 

 jacketted with a coil of wrought iron of the same dimensions as the steel 

 one, and the two cylinders were used in comparison with each other. 

 Much to our surprise, the cylinder with the steel jacket began to stretch 

 laterally under a pressure which produced no change in the wrought iron 

 coil. The experiment was considered conclusive against the use of steel 

 for welded coils, and no further attempt was made to use it for that 

 purpose. 



The duplicate of this steel coil was laid aside, and my attention having 

 been lately recalled to the subject I applied myself to discover the cause 

 of the inferiority displayed after undergoing the process of coiling and 

 welding. I had a test piece cut from the coil in the lengthways direction 

 of the bar of which it was made, and I found the elastic limit was only 

 12'5 tons per square inch, against 24 tons in the previous tempered test 

 piece ; while the breaking point was 19'1 tons, against 28'6 tons in the 

 former case. The loss of ductility was still more decided, the elongation 

 being only 7"5 per cent, instead of 36 per cent. I then had a test piece 

 cut across the welds, and this broke, not at a weld, but through the 

 solid, showing that the welding was perfect. In this case the elastic limit 

 was 12 tons per square inch, the breaking point 20"1 tons, and the elon- 

 gation 6 per cent. To determine whether the deterioration which the 

 material had sustained was permanent, or whether the quality could be 

 restored, a portion of the welded coil was hammered out in length, and 

 reduced from a piece about 5 inches broad and 2 inches thick to a section 

 of 1 inch square. A test piece from this bar showed a complete restora- 

 tion of the fine qualities of the steel and a great accession of strength. 

 The limit of elasticity rose to 21 tons, the breaking point to 27 tons, and 

 the elongation to 86'5 per cent. It was remarkable, however, that after 

 this treatment no further increase of strength was obtained by a renewal 

 of the tempering process. The fracture of the test piece from the original 

 bar was slate-coloured, and of the character usually called fibrous. The 

 test pieces from the coil showed a coarsely granular fracture, but in the 

 restored state effected by hammering, the fracture again became slate- 

 coloured and fibrous. 



Thinking it possible that the coil might have been overheated in the 

 welding process, I had a pile made with a number of small slabs of the 

 restored material, and welded at a somewhat lower heat than had been 

 jipplied in making the coil, but test pieces cut across the pile invariably 

 failed at the junctions with a very insignificant strain, showing that the 

 welding heat could not be reduced consistently with sound welds. 



In order to ascertain whether it was the heating or the hammering 

 that had injured the welded coil, I had a piece of the material cut from 

 the coil, and restored to a good condition by drawing under the hammer, 

 and then heated up to the welding point, and allowed to cool without 

 being hammered for welding. In this case the fracture showed no change 

 of crystalline structure, nor was there any decided alteration in quality 

 except that the hardening efiect of the hammering was removed. It 

 began to stretch at a low limit, namely 12 '-S tons per square inch, but its 



