melcher: change of density of sulphur 433 



The density was then determined by the loss of weight in water 

 before and after bending the sample, with the following results: 



The density of the sulphur, 1.9, was computed directly from 

 the measurements of the tubes. This value is probably a little 

 low as the diameter of the plug was taken for the diameter of 

 the tube. However, a difference of 0.1 in density only gives a 

 difference of about 0.002 in the diminution in density divided by 

 the initial density. Other errors in the experiment may have 

 arisen from wrinkles which sometimes form in the process of 

 bending of the specimen, and to the plugs which may slightly 

 protrude outside the tube unless the pin^and plug are closely 

 fitted and well soldered. The specimens were again weighed in 

 air after the experiment, to see whether there was any change 

 in weight due to leakage. 



Specunens III and VI were bent a second time with a change 

 in the percentage decrease of density from 5.4 to 5.65 and 5.4 

 to 5.47 respectively. The above data seem to indicate that as 

 the deformation approaches a complete circle the change in the 

 percentage decrease of density becomes less for equal increments 

 of strain, as specimens III and VI were bent about the same 

 amount, but VI was more nearly a complete circle than III before 

 the second bending took place. 



The percentage-decrease in density of specimen V is higher 

 than the others. The high value of V is probably due to better 

 mechanical construction and better filling of the tube than for 

 the other specimens and is perhaps more significant since none 

 of the tubes can have been absolutely free from air bubbles. 

 This specimen was bent very nearly into the form of a com- 

 plete circle and no wrinkles or protrusion of the plugs could 

 be detected. Upon opening the tube the filling was found to 

 be apparently complete throughout with no noticeable central 

 void or depression. 



The data obtained are not of the utmost precision, but are 

 close enough to show that there is a decrease in density due 

 to rupture and that the decrease in density approaches a 

 value which is of the order of magnitude of 0.0673. This is 

 the square of 0.2595 or of the interstitial space in closely piled 

 spheres of equal radius. The approximation is apparently due 



