1887.] Its Reduction to Sea Level. 261 



unnoticed, I fail to see that the formula of Leyst will be of greater 

 value than that of La Place. As I said before, the conditions are 

 totally different on elevated table lands, and it is principally to these 

 I shall devote my paper. 



No formula will ever be reliable for mountain stations higher than 

 2,000 feet, and for these La Place's as corrected is, and always will be, 

 the best. 



In reducing to a datum, it is not absolutely necessary to use the sea 

 level, almost any horizontal plane can be used, and the readings 

 reduced either up or down to it. 



But there is a great difference in the results to be obtained by the tw^o 

 modes of reduction. By reducing up we obtain a chart cutting through 

 various conditions of atmosphere, owing to the datum plane being at 

 varying heights above the ground ; at some points it will touch the 

 ground, and at others, be at varying distances above, 500 feet, 600 

 feet, or 5,000 feet or more. 



Consider the varying conditions of the atmosphere at these different 

 heights. 



I mentioned in a former paper that the storm-bearing stratum of 

 the atmosphere flows up and over the land where the slope is gradual, 

 but flows round mountain peaks ; so that over the Colony, the height 

 of this stratum would be approximately the same at Cape Town and 

 at Kimberley. Now consider this stratum to be 3,000 feet thick, and 

 let us assume it is very dense, while above flows a current of warm 

 air. A chart plotted on the plane of Kimberley would not represent 

 the actual condition of the atmosphere if reduced up from surface 

 temperatures. 



Upon a chart reduced to sea level, the pressure at Kimberley is 

 least, consequently the gradient is from Cape Town and the wind is 

 south-west. 



But by reducing up to the plane of Kimberley 4,042 feet above sea 

 level, using the surface temperature, minus the diminution for 2,021 

 feet, the reduced pressure would be much too low, the gradient being 

 from Kimberley, so that the wind would be shewn to flow to a highu 

 area instead of to a low. This will be better understood if we 

 consider the temperature at Cape Town and Kimberley to be 60^. 

 The correction reduced down for Kimberley would be with a variable 

 of 67^, but for Cape Town reduced up it would be 53^ or a difference 

 of 14*^, causing a difference in pressure of '090 of an inch. Thus if by 

 reducing down Cape Town reads 30*05 inches and Kimberley 30*00 



