OP THE UNITED STATES. ;269 



which they are taken, from the circumstance that whenever 

 they are worked, eitlier by tilinji; or rul>l)ing;. to brills'; them 

 to the proper length, they acquire unavoi'lal)lv a certain de- 

 gree of heat occasioninii; an expansion wbicli docs not sub- 

 side fully, l)efore the comparisons vvhich direct the standard- 

 in* are liiiished. 'I'he metre bcini;; therefore rendered e(|ual 

 to the oriiiinal under these circumstances, will be found too 

 short, when it is completely cooled down. Thus the me- 

 tres made l)y Lenoir are bodi shorter than the orii^iital metre 

 of the Committee of Weip;hts and Measures. It is fiom 

 this consideration that I did not l>rino; the bar fully to the 

 measure, but it still remained consideral)ly above the |)ropcr 

 len^h. Being however engaged in the comparison, I dis- 

 liked changing it, and thereby overthrew a part of my work, 

 wishing to delay it for a future time. 



On the 2tst of March, I took the dilferent standards of 

 the toise under comparison. 



The toise of Canivet being half an inch French in thick- 

 ness, and the English brass scale half an inch English mea- 

 sure, the microscopes svere adjusted to lit this toise without 

 parallax, and then the difference was compensated by laying 

 four thicknesses of white paper strips under tlie wliole 

 lenscth of the scale. The other toises had strips of proper 

 thicknesses laid under them to bring them to the same fo- 

 cus. 



'I'he distance of 7fi",8, which is nearest to the length of 

 the toise, was taken I)ctween the microscopes, from +i" to 

 78".8 on the scale, placing them nearly at equal distances 

 from both ends. 



The value of the micrometer was determined by lepeated 

 measuiements of the decimal on the scale between 7>^",7 

 and /«",><, which decimal was found to measure o", 10053 

 by the micrometer. 



I intended, as l)efore, to determine their value by other 

 intervals also ; but this being deferred until tlic end of the 

 operation, as well as the determination of the value of 

 the distance used in mean distance of the scale, neither of 



VOL. II. — N 2 



