loo American Quarterly Microscopical Journal. 



standard centimeter, which has, for some years, been adopted 

 as a standard, solely upon the reputation of the maker. Since 

 I found the two centimeters almost identical in their entire 

 length, I too hastily adopted the conclusion that my own 

 screw must be nearly correct. It turns out that neither is correct 

 at 69° Fahrenheit. They are both too long at the temperature 

 at which the U. S. Coast Survey meter is standard. The Brun- 

 ner plate is nearly free from periodic errors, but the accidental 

 errors of subdivision are large. In another standard by Brun- 

 ner, I found the first mm. to differ from the last mm. by one 

 twelve-hundredth of a centimeter. 



Of the transfers from precision screws, of course the mag- 

 nificent graduations of Rutherford stand easily first. It is 

 utterly impossible to detect any errors in the single spaces of 

 his diffraction gratings, but it is not difficult to measure the 

 residual accumulative errors from his screws. After him, and 

 since early in 1878, far ahead of all others, stand Brown & 

 Sharpe, of Providence, in point of accuracy in subdivision. 

 Of foreign standards, the disagreement between the standards 

 of Brunner and Froment is very marked. The value of the 

 centimeter which I have obtained, corresponds more nearly 

 with that by Froment. I ought to say, however, in justice to 

 these makers, that the temperature at which the plates are 

 standard is not given, at least I have been unable to obtain it. 

 It may be useful to illustrate the degree of accuracy attainable 

 abroad in the subdivision of the entire meter. 



In the following table are given the errors of the standard 

 meter of the Central Physical Observatory of St. Petersburg 

 as determined by the Director, Professor Wild, with a 

 vertical comparator of his construction. The quotations were 

 made by Herman and Pfister, of Berne. A positive correctiom 

 indicates that the space measured is too short, a negative that 

 it is too long : 



Decimeter Divisions. 



SPACE. CORRECTION. 



1 -|-. 01 79 mm. 



2 -4-. 0106 



3 -.0091 



4 —.0041 



5 --0177 



