188 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



tension in the inner layers neutralizes the compression which existed 

 there initially and eventually equals the increasing tension of the 

 outer layers, with the result that finally the stress throughout the 

 mass of the cylinder is one of uniform tension. In an ideal con- 

 dition all parts of the cylinder would be ready to break at the same 

 time, and then the maximum possible strength would result ; in any 

 actual case it is, of course, impossible to reach this ideal, but, with 

 the cylinders subjected to a preliminary stretching, it can be ap- 

 proached much more nearly than even in a built-up gun. 



The best steel to use for the cylinder is found to be a steel which 

 will stretch considerably before breaking, but which has, at the 

 same time, a high tensile strength. The best of all was a steel made 

 in this country by the electric furnace method ; this steel is a chrome 

 vanadium steel, and has, when hardened in oil, a tensile strength 

 that may reach as high as 300,000 pounds per square inch. The 

 highest pressure that I have ever found it possible to reach in a 

 cylinder has been 40,000 atmospheres, or twice the highest pressures 

 at which I have made accurate measurements. 



In the preliminary work on steel cylinders many cylinders were 

 broken. This gave opportunity for interesting observations on the 

 manner of rupture at high pressures, and two facts not to be ex- 

 pected according to ordinary theories were noted. The first was 

 the enormous amount of stretch that the steel at the inner layer of 

 a cylinder will support without rupture ; this is well shown in plate 1, 

 figures 1, 2, and 3. In the first figure the cylinder was originally 

 one-half inch in diameter, but it stretched to one and one-fifth 

 inches before breaking. The second observation was that in all 

 the cylinders tested the break started at the outside, where the stress 

 and the strain are both least; this was observed in all the steels 

 used. There is reason to believe, however, that very brittle sub- 

 stances like glass would break at the inside, as predicted by the 

 ordinary theory. The fact seems to be that if the substance is brit- 

 tle it will break at the inside first, but that if it is at all plastic it 

 will break at the outside first, the crack traveling into the inside. 



In addition to the data obtained regarding the manner in which 

 materials break at high pressures, many other peculiar facts were 

 noted during these preliminary tests. Perhaps the most interesting 

 of these is the increase in rigidity experienced in substances ordi- 

 narily soft and pliable. A striking example of this is afforded in 

 the case of paraffin, which under pressures as high as 20,000 atmos- 

 pheres becomes more rigid than soft steel, so that if paraffin is 

 forced to flow by the application of a very high pressure, and a 

 piece of soft steel is imbedded in it, the steel will flow with the 

 paraffin and will become distorted and twisted with the latter. Soft 

 rubber also becomes very hard and brittle under high pressure j in 



