4 *4 Major-General Roy’s Account of the 
glafs rods were, for the degrees of difference of temperature^ 
below that in which their refpedtive lengths were laid off, that 
this fmall apparent difference of two-tenths of an inch, be- 
tween the two modes of meafuring the thoufand feet, fhould 
have been 0.17938 in. to have made the two refults exactly 
agree, which is a real difference of only 0.02062 of an inch. 
Suppofmg then every thoufand feet of the bafe to have been 
meafured by the chain with the fame attention, and confe- 
quently with the fame, or nearly the fame fuccefs (and there 
finely cannot be any reafon to doubt of the practicability) we 
fhall have 27.404x0.02062 in. = 0.565m. or a defedtof fome- 
thing more than half an inch on the whole length of the 
bafe. 
* When the length of the chain was laid off, the heat was 66°f , and that of the 
glafs rods 68°. They will, therefore, only agree with each other accurately in 
thefe refpeCtive temperatures. The mean of twenty thermometers for the four 
chain lengths of the forty-fixth hypothenufe gave a heat of 6i°.6; and for the 
fix chain lengths of the forty-fifth, the mean of thirty thermometers gave S 9 °- 75 > 
The temperature of the 400 feet of glafs by the mean of forty thermometers 
was 65°. 3 ; and of the 600 feet, by the mean of fixty thermometers, it was 
6o°.8. Now, from thefe data, and the expanfions of fteel and glafs, as deter- 
mined by the pyrometer, the computation will ftund as follows . 
e j {”400 66.5 — 61.6 —4.9 x 0.03052— O.14955 
[600 66.5-59.75=6.75x0.04578 = 030901 
G]afs 1*400 68.0-65.3 =2.7 X 0.02068 =0.05584 
[600 68.0-60.8 =7.2 x 0.03102=0.22334 
The 1000 feet of fteel fhould have contracted more than 
the 1000 feet of glafs, 
Bunthe difference was found to be 
contraft. 
of 1000 
feet. 
contract, 
of 1000 
feet. 
Therefore the error of the chain in defeCt was « 0 , 02©62 X 27 » 4 ° 4 -“ 
0,565 in. or little more than half an inch on the whole bafe. 
