i6o 



POPULAR SCIENCE MONTHLY. 



upon, which yields a high grade of steel ; but in uniformity of 

 composition and elasticity it can scarcely be equal to the more 

 expensive crucible steel. This perhaps may at present be only a 

 matter of opinion. On such a point no definite and final conclu- 

 sion should be reached without a series of comparisons such as 

 can not be accomplished in a day. 



An unfortunate mishap which occurred at Watervliet in 1895 

 may have some bearing in this connection. In assembling the 

 parts of a forty- caliber twelve-inch rifle, the tube was, as usual, 

 rested vertically upon its breech end, and the heated jacket was 

 let down over it. The heating had been insufficient to secure all 



SlXTEEN-INCH GcN, MOUNTED ON COAST CARRIAGE. 



Weight of gun, seventy-one tons. 



the expansion needed, and as a result the cooling jacket gripped 

 the tube before quite reaching the final position intended. An 

 interesting problem was now presented, that of separating the 

 tube and jacket after they had become thoroughly cool, and com- 

 pleting the process which had been so unexpectedly interrupted. 

 The gun was provided with the inlet and outlet tubes such as 

 Rodman employed to secure a continuous flow of water in hollow 

 casting, and the exposed part of the tube below the edge of the 

 jacket was inclosed in a bag of asbestos cloth through which a 

 stream of cold air could be transmitted. The gun with its adher- 

 ent jacket and these adjuncts was let down into a furnace so as 

 to heat the jacket. Immediately a flow of cold water was started 

 through the tubes, and of cold air through the bag, while the 

 inclosing jacket was soon raised to a temperature estimated to be 

 1100 F., which was maintained for several hours. The experi- 

 ment proved unsuccessful. It was subsequently repeated twice 

 with slight modifications, but all in vain. To test the correctness 

 of the theory thus applied, a " dummy " was constructed, its parts 



