74 Proceedings of the Royal Society of Victoria. 



of the sea in Melbourne in the normal state of the atmos- 

 phere. This arrangement will have the disadvantage of intro- 

 ducing negative numbers, which will represent depths below 

 the level of the sea, and the quantities will diminish with 

 an increase of the argument, but on the whole, I con- 

 sider it better than the usual plan. Let now F = 94552'26 



- 63948-6 B, and F = 94552-56 -63948-6 B', and calling 

 a the height between the two places at the mean adopted 

 temperature, 60", we have a = F' — F. The values of F 

 or F' are tabulated in the first of the present tables, 

 with the heioht of the barometer for ai-OTment. If a 

 mercurial barometer has been used, the above value of a 

 should be corrected for the difference between the tem- 

 peratures of the quicksilver at the two stations. If T 

 and T' represent these temperatures at the lower and 

 upper stations respectively, the correction wiU be 2-5 (T' 



- T) ; as T' is nearly always less than T, this correction 

 will be generally negative. 



It will now be necessary to correct a for the actual 

 temperature of the air, calling this corrected value A, we 

 have — 



A = a + ^ !! - (t + 1' - 120) 

 956-7 ' 



putting 71 = 1 -f ^"^95'gTy ^" we have A. = na; and the value 



of n is given in the second table with t -|- t' as 

 argument. Generally speaking, the value determined from 

 the equation A.=:n (F'— F) will be far within the limits 

 of accuracy of the aneroid barometer ; but if it should be 

 thought desirable to apply the small corrections depending 

 on latitude, and decrease of gravity on the vertical, small 

 tables are given for the purpose. 



