OX THE CONSTRUCTOR <'F HOOPED CAV ft 47 



great number of minute fissures upon the outside, while the inside <li<l not show 



the least sign of crack or flaw. 



I should remark, that this ring amis made of very ton Norway iron; but. 



although I made several others in tin* sain way, and of common I\n<dish n 



well as of American iron, none of them broke under the strain before a disten- 

 sion of -j 1 ^ of their inside diameter; and, in all cases, the fracture commenced upoti 



the outside and worked gradually inward to the caliber. 



Another thing, worthy of all attention, was this: The end of each cylinder or 



ring showed, after welding, the thickm 9 of each of the several concentric rings 

 which it was formed; and after the distension, the greater diminution of thickm ss 

 in the inner and smaller rings was very apparent; thus showing how much gr< iter 

 was their distension or elongation, circumfereniially, than that of the rings outside 

 of them; and thus furnishing an experimental exemplification and corroboration, 

 if such corroboration were required, of the fact first geometrically demonstrated 

 by Barlow, and upon which he founded his theory explaining the weakm * of 

 cast-iron hollow cylinders when exposed to an internal pros are. Now, although 

 the fact is to be received as he has demonstrated it, yet it becomes evident 

 that the theory and formulas founded upon it must be limited, rigidly, to uninal- 

 leable bodies, and is in nowise applicable to cylinders of wrought metal, like the 

 rings or hoops experimented upon by me. For, to bring a case under the con- 

 ditions or facts supposed to operate in that theory, the fracture must begin 

 upon the inside, which is supposed to be distended, like a rod strained by a BUS- 

 pended weight But, in my experiments, not only was the innermost part of 

 the cylinder subjected to the straining force of the conical pin, tending to rupture 

 the whole thickness of the cylinder, but the inner portion of it, to a certain 

 depth outward, was placed between two opposing forces, viz. the pressure of the 

 conical pin in one direction, and the binding strength of the externa] portion of 

 the cylinder in the other. Between these two forces it was crushed or pressed 

 and extended laterally, and thus made thinner and longer, as a bar or sheet of 

 metal is under a hammer, or between the rollers of a mill. Under these con- 

 ditions it could not suffer fracture; for, to fracture a body its integrant atoms 

 or molecules must be separated; but in this case they were pressed more closely 

 together. This crushing pressure of the conical plug differed in no essential 



~*. J."W v t ucmug 



form from that produced by fired gunpowder. So the fracture commencing 

 upon the outside of the ring is similar to that made in the bursting of bronze 

 guns, which always commences upon the outside. The same fact was observed 



