TRUE SURFACES AND ACCURATE MEASUREMENTS ROBB. 23 



^when resting upon its three supports, or that his straight edge, 

 which was about straight when resting upon the surface plate, 

 bends slightly when he lifts it by the ends. A machine designed 

 -and constructed with the greatest care and skill, when subjected 

 to the necessary strain of belts, or the inertia of its own rapidly 

 moving parts, will spring so that the bearings, which have been 

 made nearly perfect in surface and alignment, are thrown slightly 

 awry, causing the strain to be borne by a decreased area, and 

 forcing the lubricant out. The skill of machine designers is 

 shown in designing a machine so that the journal and its seat 

 will spring together, thus preserving the adjustment of bearing 

 surfaces. We hear and see much in these days of ball bearings. 

 The success of the modern bicycle depends upon this ingenious 

 device, and has induced almost a mania for using hardened steel 

 balls for bearings of all kinds. Those engaged in mechanical 

 work often forget that the old and tried oil lubrication is really 

 a ball bearing composed of such perfectly formed balls as only 

 nature can produce, and with the advantage that they may be 

 replaced or renewed as often as desired with very slight trouble. 

 Steel balls are undoubtedly better suited to a bicycle than ordi- 

 nary bearings lubricated by oil, because it is impossible to get a 

 machine light enough for the purpose and have the bearing 

 surfaces large and parts sufficiently rigid- to keep the surfaces in 

 correct alignment ; but in heavy machinery where the strain on 

 bearings is so much greater, steel balls, with their small surface 

 of contact and tendency to crush, are not as suitable as the 

 minute spheres of a fluid such as oil. In this connection may 

 be mentioned a point which is also closely connected with the 

 second part of this subject, accurate measurements, viz., that it 

 is necessary in the construction of bearings of machinery to allow 

 sufficient space for the oil between the metals. If a shaft be 

 made as perfectly round and smooth as possible, and of the 

 exact size of the bushing or box in which it is to work, which 

 is also true, there will be no room for the lubricant. The writer 

 has seen this fact well illustrated by a bar which was intended 

 to carry revolving cutters in a machine for boring cylinders, 

 the bar and the bush were made as carefully as possible 



