98 American Quarterly Microscopical Journal. 



however different the methods of investigation employed, all 

 must ultimately reach precisely the same results, and the 

 several centimeters obtained must necessarily be comparable at 

 the same temperature within certain very narrow limits, for 

 there is but one particular meter-bar recognized as a standard 

 through the world; viz., the platinum bar in the archives of 

 Paris. 



It has been suggested, that since the more recent investigations 

 of Clark and Schubert have shown that the meter of the archives 

 falls short of its definition by one fifty-four-hundredth part of 

 its definition, it is probable that the new value thus found will 

 come into extensive use, but the attempt to make the meter 

 correspond to a natural unit has been deliberately abandoned, 

 and the International Metric Commission, the only authority 

 having the sanction of international law, has resolved that 

 the particular platinum bar in the archives of Paris, at 32° 

 Fahrenheit, shall be perpetuated forever as the original unit of 

 length, without regard to the doubtful questions which have 

 been raised concerning its correspondence with the natural 

 unit from which it was first constructed. 



Our problem then becomes a very direct one. First, we must 

 obtain a copy of the entire length of the meter of the archives. If 

 the comparison is made at a temperature differing from 32, we 

 must know the law of expansion of the particular meter-bar on 

 which the graduations are made. Having the entire meter, the 

 problem of its subdivision into equal parts is one in which dif- 

 ferent investigators would follow different methods, but, as I 

 have already said, all must ultimately reach practically the 

 same results. 



The method which the writer has pursued is described in the 

 article on two forms of comparators for measures of length.* 



As there seems to be considerable confusion as to what 

 properly constitutes a standard, it may be well to say that 

 every measure of length, claiming to be a standard, must con- 

 form to the following conditions : 



(a) If the unit chosen is a centimeter, then the distance be- 

 tween the end lines must be exactly one one-hundredth part of 

 the meter of the archives at 32°, and for any other temperature 

 it must be exactly one one-hundredth part of the meter of the 



* To appear in the April number of this Journal. 



