BRIDGMAN. 



ACTION OF MERCURY ON STEEL. 



329 



cut. The diagram (see Figure 2) showing the location of the crack in 



the several cylinders makes clear that the rupture does not take place 



along a flaw in an axial plane 



extending originally throughout 



the entire bar. Furthermore, 



since the manner of fracture of 



each cylinder demands that the 



flaw be in an axial plane if the 



fracture is due to a flaw, and 



since it seems improbable that 



a flaw throughout the length of 



one cylinder should not extend 



into the neighboring cylinders, 



the conclusion seems justified 



that the rupture is not due to 



a flaw. 



Now compare with this the 

 tests for the other set of six 

 cylinders from the same bar. 

 (See Table II.) 



The difference between these 

 two sets of tests is sufficiently striking. Not one of the cylinders filled 

 with a fluid other than mercury was broken during the tests, although 

 in several cases the test was terminated by an irrelevant accident. The 



Figure 2. Orientation of the crack 

 with respect to the original bar in the 

 six cylinders broken with mercury. 



TABLE II. 



