BY MEANS OF LONG STEEL RIBANDS. 33 
If r’ represent the corrections reduced to their aw pa value 
in® 
the last two equations are converted into 7’ = 
24 («+ Eee 
a - > hee oo, ee 
nly i ply ing the ten 
is the spring Satine With a view to the autaiciners ‘of andiibatly ‘the 
balance should be tested, and when in use the freedom of the 
attached to the spring must be ensured. If it touch the sides of the 
slot through which it passes the registered tension w ill be erroneous. 
g 
150° Fahrenheit. A difference of 50° temperature and 20 Ib. 
tension would cause a variation of °135 Ib. tension, which in a 
5-chain riband of the weight previously recommended would 
represent less than ‘002 link. And as the effect of increase of 
ultimatum of accuracy be desired. tp correction may be 
ein of a anit of sueeabered steel fora change in temperature 
from 32° to 212° Fahrenheit. In a table of expansions for thirteen 
kinds of steel this is the least, and there is a very considerable 
range in the amounts. ence, for measurements of the most 
refined Elasasipos, the coeflicient of expansion for each riband 
must be determined. In ordinary practice no error of any conse- 
quence will be introduced if 100115 be accepted as the coefficient 
for the range of temperature before mentioned. From this may be 
deduced the following simple but very accurate formula :— 
If 2d =the variation in temperature (unit 1° Fabr.) 
t= length of riband 
# = correction in hundredths of links 
dl 
R=— 
8 
and for a difference of 60° Fahrenheit, and a length of 10 chains, 
01 link. 
this formula is in error less than 
e ascertainment of the temperature of the riband in the field 
r, and even its approximate determination she — 
is no easy matte 
considerable experience. 
