358 Mr. Mallet on the Physical Conditions 



diagonal one, tending generally from the internal angle outwards; but if there were no 

 plane of weakness here at all, — if the metal were of the same cohesion per square inch 

 throughout all its parts, — the weakest place must have been that of least section of metal in 

 the directions exposed to pressure ; and as this is in a plane at right angles to the axis of 

 the cylinder, the sides of the latter would in such case have been torn directly across some- 

 where ; the cross section of fracture then being less in total area, than in case of a fracture 

 from the internal to the external angle at the cylinder's base, in the ratio of 1 : v2. 



The diagonal must therefore have been the weakest place — Why did it not break 

 straight through it ? The reason is obvious, when we come to consider the nature of the fluid 

 forces to which it was exposed before fracture. 



The normal or radial pressures against the interior of the curved sides of the cylinder, 

 and against the base at right angles to itself and to the former, commenced a rent at the in- 

 terior angle, — a certain amount oijiexure, however small, being produced in the metals at 

 both sides of it. This flexure, however slight and instantaneous, had necessarily the same 

 eflfect as if the fracture took place by rotation round consecutive points whose loci were 

 in circles all round the outer edge of the progressive fracture ; and as the greater motion was 

 in the base which was projected oif, so the fracture curved upwards, just as the fracture 

 described in burst guns turns ofl:'to one side, very near the outer surface. 



The irregularity of broken surfaces, and of the line of rupture, with reference to a plane 

 parallel to the base, was, no doubt, due to irregularities in the casting itself, or other 

 accidental conditions. 



Note G.— (Sect. 25.) 



Experiments, of the same character as those of Mr. Fairbairn, have been made in the 

 United States, upon the larger scale of casting guns, at various periods, for several years back, 

 some of which are detailed in a collection of Reports by officers of the American Ordnance, 

 published this year (Trilbner, London), and which has met my eye, for the first time, while 

 these sheets are passing through the press (May, 1856). These experiments on the effects 

 of remeltiug, or of prolonged continuance in fusion, are of the same inconclusive character; 

 and the few deductions made are sometimes anomalous and inconsistent. It is impossible to 

 avoid observing, that none of the experiments appear to have been devised with any preli- 

 minary guiding theory, based on just or sufficient chemical and raetallurgic knowledge. 



What, then, do such experiments on remelting of cast-iron amount to ? 



It has been well known for probably a century, that white cast-iron (No. 4 pig) has a 

 far higher ultimate cohesion than any of the gray, mottled, or dark gray varieties (Nos. 1, 2, 

 and 3, pig). It has been known, for nearly the same period, that the latter may be con- 



