Variation of Hourly Meteorological Normals at Kimberley. 177 



Town. I was unable to advance the investigation far for lack 

 of material. In 1901, however, Professor J. T. Morrison, working 

 on a better basis, quite independently, said that he had, with the 

 help of Mr. C. Stewart, made a comparison of the daily pressures at 

 Cape Town with those of Durban and Kimberley, and found that 

 while the several places showed the same changes, Cape Town was 

 almost invariably a day earlier than Durban. * Fig. 2 gives a 

 specimen of the sort of evidence upon which my own opinion was 

 founded. It is a diagram drawn by me in May, 1894, to depict the 

 relation between the simultaneous changes of pressure at Kenilworth 

 (Kimberley) and Cape Town. Durban pressures were added a few 

 months later from some observations kindly sent me by Mr. Nevill ; 

 and within the last few weeks I have been able to add those for 

 East London from observations kindly lent to me by the Harbour 

 Board. The comparative charts for other months of 1894 and 1895 

 are equally conclusive, but May, 1894, is given in particular because 

 in that month the Gape Times happened to print the daily observa- 

 tions without a break. The Kimberley pressures are plotted from 

 observations made three times a day, at 8 a.m., 2 p.m., and 8 p.m. ; 

 Cape Town from two, at 8 a.m. and noon; Durban one, at 9 a.m. 

 (Natal time) ; East London one, at 8 a.m. The diagram leaves very 

 little doubt that crests and depressions appear at Durban and Kim- 

 berley on the same day, at East London slightly earlier on the 

 average, and at Cape Town earlier still. This fact disposes of the 

 idea (if such a random guess wanted disposing of) that depressions 

 travel from west to east across South Africa. It also proves that 

 they do not come from the Indian Ocean, nor from the norih. 

 Plainly they come inland from some southerly direction, by preference 

 south-westerly. I have tried to carry an imaginary cyclone across a 

 map of South Africa, which should account for Fig. 2, and in which 

 the cyclonic winds, combined with the normal winds of Kimberley, 

 should give the resultant winds of Fig. 1. A cyclonic path from 

 about south-west to north-east accounts most satisfactorily for both 

 diagrams.! Further research may improve the method and lay 

 down the actual mean path of the centres of the depressions, and 



* See " Some Pressure and Temperature Results," &e., in Trans. S. A. Phil. 

 Soc, vol. xi., pt. 4, p. li. 



f To a certain extent bearing upon this is a remark of Sparrman : — "The 

 people at this place [Zwellendam] pretend to have observed that the wind, when 

 it blew from the south-east at the Cape, was always northerly with them; and 

 that, when it had ceased raining at the Cape, they had still slight showers at 

 Zwellendam." (Sparrman, "A Voyage to the Cape of Good Hope," second edition,. 

 1786, vol. i., p. 223). The assertion has probably some foundation in fact. 



