196 



APPENDIX X. 



[Revised Extract from Minute Paper No. 8 ^, on the subject of Diminished Rainfall in 

 Victoria County attached to Forest Reporl by direction of His Excellency.^ 



HONOURABLE COLONIAL SECRETARY. 



The connection between forests and rainfall in Natal is a wide subject that 

 cannot be treated exhaustively without special study and more data than are at 

 present available. The rainfall fluctuates from year to year in an apparently irregular 

 manner, but in reality it is governed by cycles with regular periods. My friend Mr. 

 D. E. Hutchins, who has made a study of the variation of the weather of South Africa in 

 cycles, finds a close connection between solar energy, as indicated by Sunspots, and 

 the rainfall.* He also traces secondary cycles, one with a period of 9'5 years, which he 

 terms the " Storm Cycle," from its association with violent gales and numerous shipwrecks, 

 the other with a period of 12'5 years, which he terms the" Mitigation Cycle." Droughts 

 prevail at the Sunspot minirnaeveryll'll years (Wolfs Sunspot period), unless masked by a 

 coinciding wave of the " Storm Cycle," or of the " Mitigation Cycle." These three 

 cycles give, in the opinion of Mr. Hutchins, a complete key to the South African weather, 

 and the character of the seasons since 1884, particularly last year's severe drought, have 

 been successfully foretold by their use. The " Storm Cycle" is the most prominent 

 feature of the Capetown rainfall, while the " Mitigation Cycle " governs more powerfully 

 the rains of the Eastern Province. I have sketched the diagram attached, which seems 

 to indicate a close connection between the waves of rainfall observed by Mr. Wilkinson, at 

 Ottawa, and the three cycles discovered by Mr. Hutchins. Many years' observations will 

 be required before these cycles can be established on a firm basis, and the influence of 

 each ascertained exactly by suitable mathematical processes (harmonic analysis). 



Two sources of discrepancies tend to mask the probable correlation between various 

 cycles and the rainfall : First, the calendar year often cuts the rainy season, and late 

 rains one year may show a dry year followed by a wetter year, though the rainfall in the 

 two rainy seasons may have been nearly the same. Second, the rain gauges may not 

 measure accurately the rainfall over a large tract of country, the precipitation being 

 influenced by locality and shelter 



Though the rainfall fluctuates in waves, the diagram shows clearly a steady diminu- 

 tion of their level. Taking two equal periods of eight years, similar in character as 

 regards .the waves of rainfall, namely, those connected with the " sunspot cycle " above, 

 a diminution of mean level from 39*0 inches during 1871-78 to 3O8 inches during 1882- 

 89, is observed to hare taken place. This diminution does not appear to be due to cyclical 

 variation, because (1) Two corresponding periods have been chosen for comparison;. 

 (2) Using all the other complete South African observations, for similar periods, which 

 I have at present with me, no such severe diminution of rainfall is perceivable elsewhere: 



Mean 



Station. Period. Rainfall. Period. 



Ottawa ... 1871-78 39'0 1882-89 



Do. ... 1871-77 40-7 1882-88 



Capetown Royal Observatory... 1871-77 26-8 1882-88 



Graaff-Reinet ... 1871-77 16'1 1882-88 



Grahamstown ... 1871-77 28-2 1882-88 



Aliwal ... 1871-77 26-0 1882-88 



Maritzburg ... 1871-77 38'4 1882-88 

 (using Gardenscliffe observa- 

 tions for some of the years.) 



D. E. Hutchins. Cyclct of Drought and Good Seasons in South, Africa, Wynberg, 1889. 



