404 



TRANSACTIONS OF THE AMERICAN INSTITUTE. 



strain from the expansive force or direct pressure of the powder, 'bending 



the staves outward; and page 47 

 jof the same book, by diagram, the 



^direction of fracture due to such 



1 strain, not through the breech, but 



'running at an angle to the plane of 



\-i- ^ 



J 



, rio:7 



3^ 



C 



FitjM 



bgOJ. 



12L 



ihe bore. 



To show that it is improbable that the direct pressure of the powder 

 should be the cause of fracture, as exhibited by the gun actually broken by 

 firing, prepare three plates of metal, say four inches thick, twelve inches 

 jPi3.9. wide, and sixty inches long, with plane surfaces; 



~ IIIZ3 ^^^^ niiddle one on being heated to 1,600° will 

 ^1 bef lund expanded one-sixtieth part of its length, 

 -■ — ' or will be sixty-one inches long. On placing 



J it between the other two, a part of its heat is 



immediately communicated to their contiguous surfaces only. Tiie expan- 

 sion of one surface of the outside plates, while the other surfaces remain 

 cold, warps the latter to the form of a segment of a circle. Now, sup- 

 posing them I laced upon the dia- 

 gram of a burst gun, the center 

 metal of which has been heated by 

 the combustion of powder, it is 

 evident that the fracture in the 

 particular direction exhibited must 

 have resulted from the unequal expansion of the gun by heat, and a dia- 

 gram exhibiting these curves, the 

 result of this expansion, will be- 

 exactly the opposite of the curves 

 on the diagram by Rodman, and 

 will account for the breaking of 

 the gun through the breech, beyond the range of the pressure made by the 

 powder. 



The following diagrams exhibit the effects of expansion of the inner 



metal by wedges. The drawing 

 4 exhibits a section of the metal of a 

 gun, with dovetail notches cut 

 along the surface of the bore. Upon 

 dri\ing wedges into the notches the 

 muzzle would be expanded, as 

 If a band were put upon the muzzle, the frac- 

 ture nearest the muzzle, and the one 

 through the cascabel, would be most 

 likely to occur first. If the band 

 were placed over the first men- 

 tioned fracture, and the wedges 

 along the re-enforce and at the 

 bottom of the bore driven most, as 

 the heat is most intense at the bottom of the bore, cross fractures of the 



r=nC 



t^hown by the dotted lines. 



