336 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



crack was in a radial plane; surrounding this crack on either side 

 was a band within which the coarser structure and the silver luster 

 showed that the steel had been amalgamated. Besides this band of 

 amalgamation flanking the crack, which was present in every cylinder, 

 there were other irregular splotches of amalgamation growing either 

 from the central hole or from the flanking band. No cases were found 

 of isolated islands of amalgamation in the midst of untouched metiil. 



The diagram gives an idea of 

 typical forms of this amalga- 

 mation, although so man}^ vari- 

 eties were seldom found in one 

 specimen. In one specimen the 

 amalgamation had grown in the 

 form of a nearly complete ring 

 with no contact with the inte- 

 rior except through the infected 

 region about the crack, in the 

 manner shown in the drawing. 

 The tool steel cylinders showed 

 the same effect, except that the 

 amalgamated band about the 

 crack was not so broad as for 

 the nickel steel, and the other 

 patches of amalgamation were 

 less numerous. The peculiar 

 fracture indicating amalgama- 

 tion was found in none of the cylinders which were broken in the 

 absence of mercury. Several such cylinders were examined, some of 

 hard or soft tool steel, and some of nickel steel, broken with water after 

 the repeated action of the maximum pressure of 24000 atmos. 



Such are the experimental facts which must be explained. It seems 

 evident that the premature breaking of the cylinders filled with mercury 

 was due to the weakening of the steel produced by amalgamation. 

 The fact of amalgamation seems sufficiently proved by the study of the 

 fracture. That amalgamated steel would be weaker than the untouched 

 steel seems obvious enough without the necessity of special experiment 

 to prove it. The fact was proved incidentally several times, however, 

 when parts of the steel packing appliances which had been used with 

 the mercury cylinders were used over again with the cylinders filled 

 with water. The packing plugs in these circumstances always broke at 

 a })ressure very much less than the normal breaking pressure and with a 

 fracture showing bright amalgamation. In explanation of the amal- 



FiGURE 9. The form of the amalga- 

 mated region in cylinders like those of 

 Figure 1 when broken with mercury. 



