lo.— THE DETERMINATION OF MEAN RESULTS FROM 

 METEOROLOGICAL OBSERVATIONS MADE AT 

 SECOND-ORDER STATIONS ON THE TABLE- 

 LAND OF SOUTH AFRICA. 



Bv J. R. SuTTOX. M.A.Cantab. 



For the purposes of this paper a daily " mean result " is under- 

 stood to be the average of twenty-four hourly observations. It 

 differs little from the true mean obtained by a planimeter from an 

 automatic continuous record. 



The ordinary time of the observations made by voluntary 

 observers for the Meteorological Commission is at present VIII. 

 Cape Colony mean time for the meridian of 22^° E., but there 

 are occasional registers where other hours have been used. It will 

 be found that it is not possible to compare together the climates of 

 two stations when different hours of observation are used at each, 

 until the departure from the true mean is determined. The station 

 using the later morning hour of the two will, on account of the 

 .single observation, indicate a climate hotter and drier than the other. 

 There is a further point, that even where many stations use the same 

 hour (VIII. C.C.M.T., say), the eastern stations shew hotter and 

 drier than the western, because of the difference between civil and 

 apparent time. Thus, e.g., there is a difference of apparent time 

 between Port Nolloth and Umtata of upwards of three-quarters of an 

 hour; and it happens, consequently, that the VIII. observations at 

 the former place are not far from 7.40 a.m.. and the latter not far 

 from 8.30 a.m., local time. Since both temperature and humidity 

 are var)ing most rapidly on the table-land between these times, 

 it follows that the results from each are not comparable until some 

 allowance has been made for the variation. Of course, in di.scussing 

 the question of mean results, the registrations of maximum and 

 minimum temperatures, which are, in a way, independent of time, 

 get over one aspect of the difificulty of comparison so far as the 

 temperature alone is concerned. The dew-points, humidity-ratios, 

 and barometric-pressures, howe^•er, are not so relieved. The object 

 of this communication is to give material for reducing these, as 

 well as the temperatures, to a common standard of reference. It 

 must be understood as dealing with the central table-land of 

 South Africa, including such stations as Aliwal North, Philippolis, 

 Bloemfontein, Hanover; and generally with all such as shew an 

 approach to equality in the ratios o ^ : «„, of the amplitudes of the 

 first harmonic terms in the formulae for the daily maximum and 

 minimum temperatures. Umtata. Queenstown. Worcester, etc.. are 

 rather outside its scope, although it will be applical)le to some 

 Eastern Province stations also in a limited degree. Stone, when 

 H.M. Astronomer at the Ca})e. collected some materials which he 



