168 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1940 



in any orientation one to the other show uniformly straight interfer- 

 ence fringes. The plane surfaces are actually developed by working 

 these three surfaces one against another in rotation, up to the final 

 stages of polishing. 



The standard plane surface of the Bureau is maintained by means 

 of three fused-quartz disks, each 28 cm. in diameter and about 4 cm. 

 thick. One surface of each is a true plane within one-hundredth 

 of a fringe. In other words, these surfaces depart less from true 

 planes than they would if they conformed strictly to the curvature 

 of the earth. 



In testing these planes, care must be used to support the disks in 

 such a way as to prevent them from bending under their own weight. 

 The stability of the fused-quartz disks has been most gratif^ang; 

 no measurable deviation from planeness has taken place in the last 

 10 years. 



ANGULAR MEASURE 



Angular quantities must be measured with great precision in cari*y- 

 ing out primary geodetic surveys. In determining the errors of a 

 completely graduated circle, the Bureau uses a special comparator 

 which is provided with four fixed micrometer miscoscopes spaced 

 90° apart around the central rotating table carrying the circle. No 

 standard circle is needed in this case because we are dealing with a 

 closed system. Errors as small as 0.2 second in a 9-inch graduated 

 circle can be measured if the graduation lines are of the highest 

 quality. 



The Bureau's circular dividing engine is used mainly for gradu- 

 ating precision theodolite circles for the United States Coast and 

 Geodetic Survey. B, L. Page has gi*aduated solid silver circles 

 9 inches in diameter to 5 minutes of arc with no error throughout 

 the circle as great as 2 seconds of arc ; that is, with no line displaced 

 from its correct position by more than 1 micron. 



THE STANDARD OF MASS 



The national standard of mass is represented by the cylinder 

 known as the prototype kilogram No. 20. It is made of the same 

 plantinum-iridium alloy as that used in the prototype meter-bars. 

 This national standard was recently taken to Paris for a new com- 

 parison with the international prototype. Its certified mass was 

 0.99999998 kilogram, a change in mass of only 2 parts in 

 100 million in 50 years. This difference is within the uncertainty 

 of measurement. 



Two other standard kilograms, one of platinum-iridium and one 

 f)i pure platinum, are used as working standards, 



