LINEAR MEASUREMENT. 



225 



comparison of line-measures with end-measures, and could easily be 

 shown if time allowed. 



It is also of the utmost importance to determine with precision the 

 rate of dilatution by heat of a standard bar, and such determination 

 can be practically made more easily and accurately from observations 

 of the defining lines of a standard bar. 



It may be desirable here to examine the extent of alteration of a 

 measure of length arising from variations of temperature and to show 

 it in the following tabular form, in regard to the materials of which 

 standards of length are ordinarily made : 



It may here be seen that an iron standard yard alters its length 

 sensibly that is to say, one ten-thousandth of an inch, with a variation 

 of temperature of half a degree Fahrenheit. At the same time, an 

 iron guage one inch in diameter will not alter the length of its diameter 

 to the extent of one ten-thousandth of an inch from dilatution with 

 less than an increase of 36 Fahr. 



The conclusion to be drawn from a consideration of all these circum- 

 stances is, that whilst for practical uses, where mechanical accuracy is 

 required, end-standard measures may be satisfactorily used, line- 

 standards are preferable for primary standards where the highest 

 scientific accuracy is needed. 



The CHAIRMAN : I am afraid we cannot go into a full discussion of 

 this very important and intricate subject, or of the relative merits of 

 measurement a but, and measurement a trait. If the commissions 

 which have been sitting in this country and in France upon this sub- 

 ject for months have not arrived at a definite solution, it is to be feared, 

 that we, pressed as we are for time, shall not be able to do so, although 



Q 



