222 Messrs. Searle, Aldis, and Dobson on Revolving lable 



supporting the table-top, and a small clip earning a pin 

 can be fixed in any position on the other arm of the rod by a 

 set-screw. The bent rod also serves as a means of clamping 

 tl.e board to the table in the manner shown in fig. 2. 



§ 5. The first step is to make the axis of the spindle of 

 the lathe-head parallel to the edge of the board. The face- 

 plate is removed from the spindle and the tip of the pin 

 carried by the bent rod is brought into contact with the 

 vertex of one of the conical ends of the spindle, the straight 

 edge of the board being in contact with the scale. The 

 board is then removed from the table and is replaced so that 

 the other conical end of the spindle is near the pin. If, by 

 sliding the board along the scule, this vertex can be made to 

 touch the tip of the pin, the axis of the spindle is parallel to 

 the ed^e of the board. In order that this may be the case, it 

 is necessary (1) that the axis of the spindle be parallel to the 

 plane of the board, and (2) that the projection of the axis 

 upon the plane of the board be parallel to the edge of the 

 board. The clamping screw enables the lathe-head to be 

 adjusted on the board so that the second of these conditions 

 is secured. In the apparatus shown in the figures no pro- 

 vision has been made for adjusting the axis of the spindle 

 so that the angle between it and the plane of the board may 

 be zero. This angle will, however, be small if the apparatus 

 is constructed w T ith moderate care. If, instead of being- 

 zero, the angle be 6 radians, the error in the measurement of 

 the radius of curvature (r) will be r(l — cos #) or ^r&\ 

 when 6 is small, and thus the error will generally be 

 negligible. 



§ 6. After the adjustment of the lathe-head on the board 

 is complete, the scale must be adjusted on the table-top so 

 that the axis of the lathe spindle intersects the axis of the 

 table. The scale and the board are first adjusted roughly by 

 eye so that one of the conical ends of the spindle is not far 

 from the axis of the table. A fixed microscope, with a vertical 

 cross-wire, is then focussed on the vertex of the cone when 

 the axis of the spindle is approximately perpendicular to 

 that of the microscope, the image of the a ertex lying on the 

 cross-wire. The table-top is then turned through approxi- 

 mately 180 c ; if the image of the vertex does not again lie 

 on the cross-wire, the difference is halved by moving the 

 board along the scale^ and the microscope is then moved so 

 as to bring the image again to the cross-wire. The table-top 

 is now turned so that the axis of the spindle is approximately 

 parallel to that of the microscope. If the image of the vertex 

 does not lie on the cross-wire, the scale is moved at right 



