69. In the preceding table, the positive, or 

 negative errors, sliow^ what portion of the break- 

 ing weight the beam bore, more or less than 

 what would have been assigned to it by the 

 formula, with the mean value, 536 or 614, of c. 

 And we see that excepting in experipient 24, 

 where the error was rr, the positive errors never 

 amounted to above t^, nor the negative ones to 

 more than iV> and both were generally much 

 smaller. 



70. Since 636, or 614, is the mean value of 

 c, to obtain the breaking weight in cwts. by the 

 formula, according as the beam has been cast 

 erect or on its side, one twentieth of the above 

 numbers will be the value of c, to produce the 

 breaking weight in tons. 



636 



erect. 



c — %r, = 26.8 to obtain tons, in beams cast 



514 

 20" 



on their side 



c = 20" = ^^'^ *^ obtain tons, in beams cast 



