February 9, 1905] 



NA TURE 



343 



Cape Colony is subject to another type of weather, due to 

 the passage of storms from the South Atlantic, the " roaring 

 forties " of mariners. It is necessary very carefully to 

 distinguish between these two weather systems. In the 

 one the storms bring winter rains to a small part of the 

 subcontinent, i.e. Cape Town and the south-west ; in the 

 other the storms precipitate the abundant moisture brought 

 by the trade winds from the Indian Ocean, more or less 

 over the whole subcontinent. 



This much of e.xplanation is necessary in order to under- 

 stand clearly the connection between the weather of India 

 and that of South Africa. In studying this connection we 

 have at the outset to eliminate the winter weather of the 

 south-west with its winter rains coming from the South 

 Atlantic. 



Sir John Eliot, in his reply to Mr. Sutton, very properly 

 excludes the area of winter rains. I go further, and 

 exclude what Mr. Sutton has termed the area of spring and 

 autumn rains. The latter are areas where, with the winter 

 storms still prevailing and the summer south-easters com- 

 ing in from the Indian Ocean, there is the most marked 

 precipitation in spring and autumn. We are not in a 

 position to say how far these rains have been produced by 

 the tail-end of the retreating Atlantic storms or by the 

 head of the advancing humid south-east trade currents. 

 The fertile country watered directly by the south-east trade 

 is comprised in sections x. to xv. of Mr. Sutton's rain- 

 fall areas, viz. the east of Cape Colony, Kaffraria, Basuto- 

 land, the Orange River Colony and Natal, and, in addition, 

 all the Transvaal, Rhodesia, and the Portuguese territory ; 

 in fact, it is the whole of fertile South Africa with the 

 exception of the southern and south-west coasts. In the 

 table below I give the mean of Mr. Sutton's figures for 

 his sections x. to xv., comprising Eastern Cape Colony. 

 Transkei, Basutoland, Orangia, and Natal, and I add 

 the yearly rainfall from typical stations in the Transvaal 

 and Rhodesia, as correct general average figures for these 

 territories are not available.- 



Percentages of Rainfall in the ■Stimnier Rainfall Areas, 

 1S91 to 1902 : Mean of Sutton's Sections x. to xv. 



1891 1892 1893 1894 iSg5 1896 1897 i8g8 1899 1900 1901 1902 

 Per cent. 

 136 106 132 97 103 102 74 107 89 82 98 93 



And correcting Sir John Eliot's table to purely summer 

 rainfalls it will read thus : — 



These figures show more strikingly than those already 

 quoted by Sir John Eliot the intimate connection between 

 the rainfall of India and South Africa during the period 

 1S92 to 1902, and the connection would have shown better 

 if seasonal instead of calendar years had been taken, since 

 the calendar year cuts into two unequal portions the South 

 ."Xfrican summer rainfall. It will be noted that each Indian 

 famine year has been followed by one or two particularly 

 bad vears of drought in South Africa. 



It is a somewhat remarkable coincidence that, while the 

 number of Nature containing this discussion was on the 

 sea on its way to the Cape, I prepared my yearly forecast 

 of South African weather, and in that took occasion to 

 point out the very close connection of the two rainfalls 

 during this period. I may perhaps crave your indulgence 

 to reproduce it, since it confirms so singularly Sir John 

 Eliot's view. Speaking of certain typical stations I said : — ■ 



" Sir John Eliot's paper shows that 1892, 1893, and 1894 

 were years of good rainfall in India. These were the last 



years of general good rainfall we had in South Africa, 

 in 1895 'l^s drought set in at most South African stations. 

 P'urther, in this droughty period there were two years of 

 bad famine, viz., 1S96 and 1S99. These two years of 

 famine in India were the two worst years of drought at 

 many typical South African stations. At present we are 

 not in a position to obtain average figures for the whole 

 of South Africa, but nearly the same purpose will be served 

 by taking certain typical stations thus : 



" At Bulawayo (Hope Fountain), in iSgo-i, there was 

 the heaviest rain on record, viz., 45 inches; all the fol- 

 lowing years have been years of drought except three years 

 when the rainfall was barely above the average. 



" At Johannesburg there were good rains in 1894, when 

 there were good rains in India, fair rains in 1895, and 

 then drought, when there was drought in India. 1896 

 (one of the Indian famine years) was the worst year of 

 drought in Johannesburg. The great Indian famine^ of 

 1S99 was represented by a bad drought 1898-g preceding 

 the failure of the Indian monsoon by four months. . . . 



" Natal rainfalls correspond closely %vith the Indian 

 rainfalls. While 1S99 was the worst famine for many 

 years in India, 1S99 and 1900 were the two worst years 

 of drought ever experienced at Durban, in Natal, since 

 meteorological observations were begun there in 1866. In 

 1900, the Durban rainfall was only 27 inches against an 

 average of 41 inches. .At Maritzburg, representing the 

 inland Natal districts, 1S99 was also a year of drought, 

 but the greatest deficiency was registered the following 

 year (probably chiefly due to the calendar year dividing the 

 seasonal year). 



" Again, at Grahamstown, Cape Colony, in 1899 there 

 was under 20 inches against an average of 29 inches ; 

 at King William's Town in 1899, only 16 inches against 

 an average of 25 inches ; while at Graaff-Reinet in 1899. 

 there was only 9 inches against an average of 15 inches. 

 At all these South African stations, 1899, the great Indian 

 famine year, was the worst year of drought in recent 

 times ! " 



The rainfall curves for Uintata, Evelyn Valley, and 

 Katberg show similar features, viz., severe South .African 

 droughts corresponding to the years of Indian famine, and 

 a general deficiency of rainfall corresponding with the 

 years of general deficiency of Indian rainfall. The rainfall 

 curve of Evelyn Valley (Fig. i), however, is very remark- 

 able. This is a forest station, and the observer a par- 

 ticularly good one. I have elsewhere compared this 

 station to Cherapunji, in India. I founded this station in 

 1887, and it has since shown the heaviest rainfall on the 

 summer register. It lies in a cul de sac of the mountains 

 facing the south-east at an elevation of 4200 feet. I have 

 long regarded it as the typical southern station for the 

 summer rainfalls. A study of its yearly rainfall curve 

 shows how rain failed here in the most striking manner 

 previous to the Indian famine of 1896, and during and 

 after the Indian famine of iSgg. 



With regard to Mr. Sutton's statement that there has 

 been no severe drought during recent years in South 

 Africa, there is abundant evidence to the contrary. 



A year ago I wrote : " In the Karoo the present drought 

 is considered the worst during the last half-century. At 

 Hanover (Upper Karoo) during nearly a year there has 

 fallen only three-quarters of an inch, the normal yearly 

 rainfall being 15 inches. The drought has lasted on and 

 off since 1896-8, and during the worst years cattle and 

 sheep have perished in millions. In British Central Africa 

 the drought has lasted since about 1898 ; it is reported that 

 the Shir(5 Lake is now nearly dry. Last summer's crops 

 in the Transvaal, so sorely needed after the war, were a 

 complete failure, while in Natal, Rhodesia, and the country 

 to the north there was in many places famine, and people- 

 dying in places too remote to 'be reached by Government 

 aid. 



" When will the drought end? is now the great question 

 for the countrv. 



" Good rains have fallen recently all down the eastern 

 side and on the south coast of South Africa. This rain 

 has come as a precious mitigation of the drought. It may 

 be looked on as a favourable indication for a good season 

 — perhaps more favourable if it had come later. 



" The local and other indications of an early ending of 



NO. I 84 I, VOL. 71] 



