404 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



posed to a test of 2000 pounds hydraulic pressure, and were 

 supposed to be sufficiently strong for the strains that they 

 would be called upon to bear. The flasks having burst, how- 

 ever, in the course of the experiments of the first day, it be- 

 came necessary to thoroughly examine the methods of con- 

 struction. The holders, as made by Mr. Lay, were, it seems, 

 at the higher temperatures of the contained gas, subjected to 

 pressures such that the strain on the iron was more than one 

 third, and sometimes one half the possible strength of the 

 material. The chances of flaws in the metal, and the devel- 

 opment of weakness under strain, are too great to allow of 

 working so close to the limit. Soft iron, in fact, is not a 

 good material for the construction of flasks to be submitted 

 to great strains, which often increase suddenly. Two meth- 

 ods were proposed in making new flasks; one of which was 

 the construction of a flask of copper deposited by electricity, 

 thus getting a vessel without joints. The other plan con- 

 templated the use of soft iron, but by a mode of construction 

 which seems to be much better than the one employed by 

 Mr. Lay. Subsequently a new and superior process was de- 

 vised and adopted to the exclusion of all others, and this 

 was the result of the combined skill of Messrs. Matthews 

 and Hill. Experiments had to be made in order to find out 

 the best methods of manufacture ; but eventually flasks were 

 made to contain liquid acid which proved entirely satisfac- 

 tory. These are cylinders with round heads, each cylinder 

 being provided with one valve in the centre of one head. 

 When in place in the torpedo, the cylinders lie upon their 

 sides, and a tube leads from the opening controlled by the 

 valve into the interior of the cylinder, being there turned up 

 against the upper side, and in this way only the gas can pass 

 out when the valve is opened. Each torpedo is provided 

 with four cylinders, two of them seven feet long, one five, 

 and the other four feet long. The outside diameter of each 

 is twelve inches. They are made of the finest sheet steel, 

 nearly one twentieth of an inch thick, in successive layers 

 which are firmly fastened together with pure tin. The cyl- 

 indrical portions of the flasks are made by rolling up a sheet 

 of steel, of the proper length, into a cylinder. Another sheet, 

 somewhat longer, is rolled into a similar cylinder, which is 

 slipped half over the length of the first. A third, fourth, etc., 



