82 
BALL-BEARING FOR ONE INCH SHAFT— CONTINUED. 
(Oil Used.) 
Weight | Weight. Rev. 
| 
Maes sembeder pounds, «| Smindibes te avonds {aoe ee 
a ee 180, in eg a0 bones ee 0018 
eee ede tage i aa tee: a; one ets 0018 
2 ee le Nas | 768 BD: = | Mone! esecee 0018 
Three sets were made for each load, and, with the weights used, the 
coefficient of friction varied from .0017 to .0022, as averages for the three 
sets. Of course, the width of this range may be due to inaccuracy in 
reading from the scale, as the variation in pull on the scale arm caused 
a rapid vibration of the scale index. 
The bearings used were those supplied on the market for carrying 
shafts, and the principal cause of the jar in the apparatus during the test 
was due to slight inaccuracies in grinding the races. 
In another set of tests, where the load was increased to seven hundred 
pounds, it was found that somewhere between the six hundred and the 
seven hundred pounds load the balls and races had become pitted, small 
pieces of the hardened steel being torn from the surfaces. These pieces 
were found in the race-way or in the oil that was used. It was found, 
further, that the tendency to heat was much reduced when oil was used 
and that the whole movement was smoother and steadier. 
FurRTHER STUDIES IN THE PROPAGATION OF SounD. By A. WiuMER DUFF. 
{Abstract.] 
In a previous paper the writer gave a theoretical discussion of the 
propagation of sound in spherical waves, allowing for the effect of the 
viscosity of the air and the conduction and radiation of heat from the 
condensations and to the rarefactions. It resulted from this investiga- 
