1884.] Mr. Gamble, Meteorological Notes, 13 



Dr. H. Leach, whose heights are all much too high is in this case seven 

 per cent, out, 4700. The readings of the mercurial barometer, carried 

 by the Secretary of the Meteorological Commission on one of his inspec- 

 tions, gave 4023 or nine per cent, too low. 



The private meteorological observatory at Aliwal North is 4330 

 above sea level by spirit-levelling. By comparing the mean barometric 

 readings of the years 1877, 1878, 1879, with the mean at Port Elizabeth 

 for the same time, I make 4375 or one per cent, too high. By com- 

 paring '77 only with the Eoyal Observatory, I make 4326, '78, 4337, 

 '79, 4368, '80, 4334, which are remarkably close. 

 The height of Pretoria is given somewhat variously : 



Dr. Leach . . . . . . 4620 



Jeppe . . . . . . 4450 



Serpa Pinto .. .. .. 4298 



A. Anderson . . . . . . 4090 



Baines .. .. .. 4007 



As also is Lydenburg : 



Baines .. .. .. 5825 



Erskine .. .. .. 4781 



Cohen .. .. ., 4706 



Sir Chas Warren . . . . 4350 



I think we may say that simultaneous barometric readings con- 

 tinued for a year with all proper precautions should give the height 

 within two per cent. When however heights are got from a single 

 observation or even from several observations taken much about the 

 same time, an error of ten per cent, may be expected. 



The great uncertainty of the air-temperature is the great drawback 

 to accurate work. The correction is a rather large one, the factor being 

 t^ -t- t^ - 64 . 

 900. 

 Curves showing the fluctuation of pressure in the yearly period have 

 been drawn for three coast stations, Poyal Observatory, Mossel Bay 

 and Port Elizabeth, and for five up-country stations, Worcester, 

 Lovedale, Aliwal North, Bloemfontein and Sutherland. [The tables 

 from which the diagrams were drawn are given below]. 



All the averages are, of course, corrected for temperature of 

 mercury, but only the observations taken close to the sea are reduced 

 to sea-level for reasons given above. 



The early appearance of the maxima and minima for the j^ear at the 

 up-country stations is noteworthy, pressure agreeing with the tempera- 

 ture in this peculiarity. 



There is at most stations a very remarkable hesitation or bend in 

 the month of May. At some stations this is not so clear, and this is 

 the case at the Eoyal Observatory. I believe this depression is 

 a true feature, and that the reason why it is not shown in the 

 Eoyal Observatory's diagram is that the periodic minimum frequently' 

 occurs there in the beginning of June instead of at the end of May, 

 thus masking or smoothing out the depression when the curve is only 

 drawn from monthly averages. I am in the hope of being able to get 

 out five day means for some places, which will no doubt throw light 

 on the question. It seems probable that this May depression is 

 analogous to the well-known '^ November wave," of the northern 

 hemisphere, and perhaps also to the short period of unusually mild 

 weather generally experienced before the beginning of winter, and 

 called ''the Indian " or St. Martin's summer. 



