306 SURVEY OF THE COAST 



To make tliese results comparable, it is necessary to re- 

 duce them all to one temperature, by the difference of ex- 

 pansion between iron and brass. I shall for this purpose 

 make use of the results of my pyrometric experiments, 

 made immediately after this comparison, aud described in 

 p. 224, Vol. I. N. S. of the Transactions of the American 

 Philosophical Society, the mean results of which gave the ex- 

 pansion expressed in decimal parts of the whole length, for 

 one degree of Fahrenheit's Scale, as follows : — 



In iron,=0,000006963535 



In brass,=0,000010509030 



Difference,=0,000003545495 



As all tlie details of the comparisons are here stated, it 

 will be easy to apply any other expansion in the further cal- 

 culation of the result, if desired. 



The temperature having increased during the comparison 

 with considerable regularity, and my work having been un- 

 interrupted and uniform from the beginning of each series 

 of comparisons to the end of it, the temperature correspond- 

 ing to each of the results may be considered equal to the 

 mean temperature between the two observed. The tempe- 

 ratures adopted in the table of results were therefore deter- 

 mined upon this principle. 



I found it best to reduce all comparisons to the tempe- 

 rature of 32° Fahrenheit, or 0° Centesimal and Reaumur, 

 for both the brass and iron, as we may easily obtain this 

 temperature in nature, and it can therefore be presented by 

 experiment ; which would not be possible, if the value of 

 the French measures at 32° were given in a length of the 

 English brass scale at 62°, since this would always introduce 

 a result of mere calculation. I shall reduce from iron to 

 brass : so that the length of the metres will be given in En- 

 glish inches in brass, at the temperature of 32° Fahrenheit. 

 The brass metre is therefore in this case considered as need- 

 ing no reduction. 



