SOME MKRCURY STANH.\i;i>s OF RESISTANCE, ETC. 



77 



at right angles to the axis, and the grinding arrangements were primarily intended 

 to serve this purpose. Afterwards, however, a slightly convex surface was chosen, 

 for reasons hereafter given. The figure (fig. 5) will best serve to explain the 

 motions. 



In the diagram (fig. 5) M is the small motor mounted on the slide rest, and, 

 therefore, capable of travelling in the two directions <<! and ;/' mutually at right 

 angles. Other directions of motion inclined to cd were also possible by varying the 

 angular reading at S. At the outset the reading at S was zero, and by attaching a 

 wire pointer to the disc D the line of travel was determined. This is indicated by 

 the line ab in the figure. At right angles to this cd was set off, and three short 

 clamping pieces A, B, C were fixed so that their grooved portions were parallel to the 



Fig. 5. 



line cd. The plane in which the disc D rotated was now at right angles to the axis 

 of the tube. D being a small disc about l inches in diameter, it was thought 

 possible, once an end of a tube was approximately plane, to make it perfectly so, and, 

 in addition, to lie at right angles to the axis. An attempt to attain this end by 

 clamping the tube somewhat rigidly while D travelled to and fro failed. The failure 

 is mainly attributable to the rigidity of the tube, for the vibration note was very 

 high, and the risk of breaking considered too great. Instead, therefore, a slightly 

 convex surface was ground. 



At the commencement, a not too fine circle of emery cloth was attached to the disc 

 and the speed adjusted to about 5 revolutions per second. The grooves in which the 



