384 Special Reports 



the U was in position on the pier a vertical plane through these marks normal to the 

 surfaces passed through the axis of the spool. The micrometers were mounted with 

 their anvils, screws, and slides horizontal and the axes of the anvils in this vertical 

 plane. 



The anvil of each micrometer was a cylinder of hardened steel, about 6 mm. in 

 diameter, and was fitted with just sufficient play to move freely in a brass block fastened 

 to the slide. Rotation was prevented by means of a pin and slot in the usual way. 

 The measuring end of the anvil was cut down symmetrically until the terminal surfaces, 

 as wide as the diameter horizontally, was only about 1/2 mm. in height. This end was 

 ground and lapped flat and normal to the axis. The other end of the anvil was rounded 

 and during a measurement pressed against a brass arm pivoted on a horizontal axis 

 normal to the anvil and carrying, with its axis in the (vertical) plane of motion, a half- 

 minute spirit level. A light steel spring forced the level holder against a stop, or, during 

 measurement, against the anvil. The force on the anvil necessary to move the level 

 holder from the stop was less than one-quarter kilogram. 



The slides and their ways were ground so flat and made so true that the zero read- 

 ings of the levels were not affected by a tenth of a division (i.e. by 1/20', corresponding 

 to about 0.2 n in the micrometer reading) when the slides moved over a number of 

 millimeters. The slides were driven by the anvils of Brown and Sharpe micrometers with 

 heads reading directly to 0.01 mm. and with estimation to 0.001 mm. easy. To facilitate 

 the settings, fight brass wheels about 6.5 cm. in diameter were attached to the barrels. 

 The slides were held against the (non-rotating) anvils by suitable springs. Backlash was 

 very minute, but was avoided in the usual way. 



Two end standards were used, and two corresponding methods of procedure. In 

 one method a Brown and Sharpe 300 mm. standard, with ends ground spherical to 

 the same diameter, was mounted, approximately central and horizontal, by means of 

 bushings, in one of the pairs of holes diametrically opposite in the spool, and the mi- 

 crometers were mounted with the axes of their measuring anvils in a horizontal plane. 

 Alternate settings could then be made upon the end standard and upon the wires at 

 the ends of any horizontal diameter. In this way the mean diameter of either double 

 coil could be determined. 



In the other method one of the micrometers was raised 1 mm., so that the mi- 

 crometers could be set upon the opposite ends of the diameters of the individual turns. 

 The standard used for this arrangement was made from a Brown and Sharpe 325 mm. 

 end standard with spherical ends ground to the same diameter. This was cut in two, 

 and one piece moved laterally 1 mm. with respect to the other and at the same time 

 the length was reduced to about 299.64 mm. The pieces were screwed and soldered 

 together with their axes parallel. At the same time suitable bushings and adjusting 

 rods were provided for the proper mounting of the standard with the axes horizontal 

 and one a millimeter higher than the other in the vertical plane. 



The length of the first standard and its temperature coefficient were determined 

 at the Bureau of Standards. The length of the second, or broken, standard, was 

 determined here with the measuring instrument described above and the first standard, 

 the coil being an intermediary; and its temperature coefficient was assumed to equal 

 that of the other standard, viz, 11 X 10" per degree C, a procedure which was entirely 

 justified. 



Most of the measurements were made by the second method, as it alone gave 

 information about the individual spirals of each pair making up a Helmholtz coil, and 

 only these measurements have been reduced. 



Most of the measurements were made at room temperatures near 23 C. Other 

 measurements were made at temperatures about 9 C. lower in order to determine the 



