Comparative Micrometric Measurements. By M. I). Ewell. 543 



with the first millimetre of an invar decimeter, the corrections 

 of which are derived from both the National Bureau of Standards 

 and the International Bureau of Weights and Measures, we have 

 the absolute correction of the first ^ mm. of "Powell and Lealand 

 No. VIII." = + 0*9 yx or + 0*45 /x on each r g mm., composing 

 the said j 1 ^ mm. After applying all corrections, the first jjJq mm. 

 of " Powell and Lealand No. VIII." therefore equals 49 '33 fi, and 

 the second T g Q mm. equals 49*77 fx. The said corrections are re- 

 spectively -r- • 22 fM + • 45 /x = + • 67 fi and - • 22 fx + • 45 fi 

 = + • 23 ix. 



The total correction of the first yo 6 mm - °* 

 Lealand No. VIII." being + 0-67 /x, the absolute 

 each j^q mm. composing this space is + - 13 fx. 

 the relative and the absolute corrections to each r J n 

 have finally : — 



" Powell and 

 corrections for 

 Applying both 



mm. space, we 



The methods adopted by the writer are quite different from 

 those adopted by Mr. Keeley. It has been the practice of the 

 writer, when the unit measured is small enough not to occupy 

 more than about one-third of the field, which, or even less, with 

 ordinary objectives is about the limit, to traverse the whole space 

 measured with the movable wire, taking the difference between the 

 initial and terminal readings, and always using the same part of 

 the screw. Practically the same results have been obtained by the 

 writer using a single web, two webs crossing at an acute angle, and 

 two parallel wires, the space between them being bisected by the 

 defining lines. The last is, in the writer's experience, much more 

 convenient than the other two, and obviates any error incident to 

 the use of one line only where the two defining lines are of unequal 

 thickness. The fixed wire is used simply to be certain that the 

 space being measured has not moved during the process of measure- 

 ment. 



Mr. Keeley's method is radically different. It will be described, 

 using his own language : — " All measurements are made between 

 the fixed and movable wires in the filar micrometer. The fixed 

 wire is placed about two turns of the screw to the right of the 

 centre of the field and adjusted by set screws, so that when the 

 movable wire is exactly over it. the drum reads zero. As it is 



