by means of the Temperature at which Water Boils. 33 

 or, as good approximations in two terms, 



£=520-48 ft. (212°-T) +0*967 (212°-T) 2 , 

 or 



£=520-18 ft. (212°-T) + (212°-T) 2 . 



But from equation (8) we also obtain 



W212= S = 115 ' 71976; 

 and in the same manner as before, 



w=115-71976-0-194757(212°-T), [. . . (9) 



and ' i <m- i t> 5-1346 (212°-T) /im 



log30m.-logB = m .J: 74A + T >- • • ( 10 > 

 and 



•••^309971 ft. x ^5=^, l (n) 



orlogA=5-491321+log(212°-T)~log(382-174 + T). J 



Or, avoiding the fraction in the divisor, we may employ the 

 formulae 



on oo oi 



A=309 880ft. x gy, 

 and 



log/j=5-491194 + log(212° + T)-log(382 + T); 

 or 



212°+ T 



£=308837 ft. x 380 + T , 



and 



log h= 5-489729+ log (212°-T) -log (380 + T), J 



(12) 



either of which will give almost identically the same results as 

 equation (11), the last of the two (owing to a very small second 

 difference in the value of n which has been neglected) being 

 theoretically the most accurate of the three. 



6. If the boiling-point is observed at two stations whose differ- 

 ence of level is required, writing d=212°-T, and d'=212°-T', 

 we have 



£'-£= 309971 (T-T') 



170-174 + T + T' + -00169^' 

 and 



log (h'-h) = 5-491321 + log (T~T')-log(l70'i74 + T + T'+ ~^; (13) 

 Phil. Mag. S. 4. Vol. 25. No. 165. Jan. 1863. D 



