264 A. G. Howoird. — The Barometer : [Sept. 28, 



sea-level temperature of 60^ the decrease is 1^ in every 571 feet. The 

 height of the limit of perpetual snow decreases towards the poles in 

 the proportion of the squares of the cosines, and the altitude wherein 

 the temperature decreases P will vary accordingly. 



By the term tropic I mean that parallel of latitude where the sun 

 shines vertical. The calculation for decrease must be made from this 

 parallel and not from the equator, bearing in mind that solar action 

 lags behind nearly a month. The decrease of temperature due to 

 altitude which I have mentioned is the mean decrease ; the nearer the 

 surface of the earth, the quicker does the change take place. This 

 question of decrease of temperature ought to be gone into practically, 

 and I should like to see it done here. We have Table Mountain and 

 Lion's Head. Why not utilize them ? A wet and dry bulb on 

 Signal Hill would help. Where "there's a will there's a way, and I 

 should like to see South Africa take ofl: the palm in this question. 



I will now pass on to some practical results. My first attempt to 

 construct charts from barometer readings at high altitudes, was early 

 last year. I compiled thirteen charts for 1885, twelve monthly means 

 and a yearly one. I used Hazen's tables of reduction, and was very 

 successful. 



On attempting to plot daily charts, I found that temperature, which 

 had been reduced to a mean in the monthly ones, seriously affected 

 4he reductions. One station had too high a reading, another too low, 

 consequent upon the height of the thermometer, and the greater the 

 altitude the greater the error. . 



I was so annoyed at this, that 1 destroyed the whole of them, and 

 plotted new ones, using only those reductions for a temperature of Q1°. 

 By these means the differences were done away with, and I came to the 

 conclusion that temperature had nothing to do with a formula for reduc- 

 tion on elevated plains. I then began to work out tables of my own. 



The result of my work I now place before you. I do not assert it 

 is perfection, nor do I say the formula is new. A new formula is not 

 wanted. As I said before, mine is to be used solely in reducing from 

 table lands, and by using it we do away with the discrepancies due to 

 the temperature variable, and with the anomaly of a temperature 

 <jorrection where no atmosphere exists. 



I have compiled tables of my own for all the principal places in the 

 Oolony from which returns are received, and have been successful so far.* 



* Charts for the first six days of January and of July, 1885, were exhibited at the 

 lueetiug. 



