NA TURE 



[November 3, 1904 



its relation to that of India. Mr. Hutchins was for some 

 years in the Madras Forest Department before he went to 

 the Cape some fifteen or twenty years ago. He has made 

 a special study of the rainfall of South Africa, and is a 

 careful and enthusiastic investigator in rainfall problems. 

 He is, from his double experience in India and .South Africa 

 and his present official work and position, eminently qualified 

 to form a judgment on the abnormal features of rainfall 

 distribution in either area, and on their economic effect. 

 It is hence, as I hope to show later, very satisfactory that 

 Mr. .Sutton's figures confirm the general inferences I made 

 about South African rainfall, based chiefly on Mr. 

 Hutchins's information, in my address. 



Before discussing Mr. Sutton's data and inferences, 

 perhaps I may be permitted to deal with two or three im- 

 portant issues raised in Mr. Sutton's letter. 



The first is contained in the opening paragraph, in which 

 he says " south-east winds are rare on the south-east coast 

 of .South .\frica, and the rain of the greater part of the 

 tableland and north-east coast comes mostly from some 

 northerly direction." If these casual remarks have any 

 point at all, I think I am correct in assuming that they 

 imply that Mr. Sutton considers the rainfall in the areas 

 mentioned is not due to humid currents from the Indian 

 Ocean, but from the dry interior to the north of the table- 

 land. 1 have examined the rainfall charts of South Africa 

 given in Bartholomew's "Meteorological Atlas," and they 

 certainly indicate to me that the aqueous vapour, the con- 

 densation of which gives rainfall in the eastern half of South 

 Africa, is brought up by air movement from the Indian 

 Ocean, and occurs as a summer precipitation. Hence, so 

 far as I can reasonably judge, that area forms a part of 

 what I have termed the Indo-oceanic region. I might add, 

 in further reply, that rain in certain parts of India during 

 the south-west monsoon chiefly occurs with easterly and 

 north-easterly, and even with northerly winds. But these 

 facts have not yet been utilised by anyone to prove that the 

 rainfall is not brought up from the adjacent seas and oceans 

 by the south-west monsoon circulation. 



-Mr. Sutton in a later paragraph says he fancies that 



the impression of unusual dryness over South Africa in 

 recent years arises from the misleading mean values used 

 by the Meteorological Commission for comparative purposes 

 which are taken from Buchan's rather futile ' Rainfall of 

 South Africa,' and average fully two inches (equal to 

 perhaps lo per cent.) too great." There is an air of 

 certainty about this statement which I am unable to share 

 without further proof. Buchan's means are based on ten 

 years' data, Mr. Sutton's on twenty years' data. It does 

 not necessarily follow that twenty years' means are better 

 representatives of normal or average conditions than ten 

 years means. It depends entirely upon whether the ten 

 years may or may not be accepted as representing the normal 

 conditions, and whether the additional ten years' data are for 

 an abnormal period or not. The fact that the two sets of 

 means differ on the average of the whole area by lo per 

 cent, indicates to an outsider on South African meteorology 

 like myself that it is quite as probable the ten years' 

 additional data erred in defect as that the ten years' data 

 employed by Dr. Buchan erred in excess. There hence 

 appears to be (in the absence of any proof) an element of 

 doubt in his means, just as he asserts to be the case in the 

 " rather futile " means of Dr. Buchan. 



.-\gain, if I read Mr. Sutton's letter rightly, he considers 

 that the question as to whether the crops have failed over 

 large areas being due to drought is settled by a consideration 

 of percentage variations. It is certainly not the case in 

 India. \ percentage variation gives no certain indication 

 unless considered in relation to the normal fall, and also to 

 its time-distribution. A deficiency of 25 per cent, is of abso- 

 lutely no economic importance iii such areas as Sind (with 

 an average rainfall of about four inches) or such as Arakan 

 (with an average of more than 200 inches). The former 

 area depends solely on irrigation for cultivation, and the 

 latter is so abundantly supplied for the rice crop that it 

 bears a loss of fifty inches lightly. On the other hand, in 

 the regions termed the dry zones in India, where the mean 

 ramfall ranges between fifteen inches and thirty inches, a 

 deficiency of 20 per cent, is usually a serious matter, more 

 especially if it accompanies more irregular distribution than 



NO. 1827, VOL. 71] 



usual unsuited to the staple crops. Local knowledge of 

 the agricultural and economic conditions is hence of the 

 greatest importance in estimating the probable effect of a 

 given variation of rainfall in any area. Mr. Hutchins, I 

 have every reason to suppose, possesses such knowledge for 

 South Africa, and hence I attach the highest value to his 

 information on such matters. 



The evidence I have collected, a small portion of which 

 was given in my address, appears to me to have established 

 that during the period 1895-1902 there was a marked 

 tendency to more or less continuous deficiency of rainfall 

 over the Indo-oceanic area, most pronounced in dry inland 

 districts, and which in India intensified into severe droughts 

 in the years 1896, 1899, and 1901, diminishing the crop 

 returns over large areas to such an extent that it was 

 necessary to resort to famine relief on a large scale during 

 the twelve months succeeding each period of crop failure. 



I was unable to make as precise statements for either 

 .■\ustralia or South Africa, but the scanty facts and inform- 

 ation at my disposal appeared to justify the statement that 

 these areas were similarly affected. I also pointed out that 

 this period stood in marked contrast to a preceding period 

 of three years, 1892-4, when the precipitation was apparently 

 in general excess over the same large area. 



I give in the following table a comparison between the 

 rainfall variations of India, and the area of spring, 

 summer, and autumn rains in South Africa, which, so far 

 as I can judge, is mainly dependent on the Indian Ocean 

 supplies of aqueous vapour. I give, in the absence of the 

 number of stations for each area, the arithmetic means of 

 the second and third horizontal rows of figures in Mr. 

 Sutton's summary of his data : — 



', Period of general ( 



Period of general deficiency of r 



1892 



India 

 + 12 

 -f22 

 + 16 



These figures show that the eastern half of South Africa 

 had heavier rain than usual during the same period (1892—4) 

 as India, that it was steadily in defect during the first 

 five years of the period of persistent deficiency of rain in 

 India, and was especially deficient in the years 1897 and 

 1S99, the former being the year and rainfall season following 

 the first severe drought year of the period in India, and the 

 latter the same year as that of the greatest drought experi- 

 enced in India during the past 100 years at least. The 

 parallelism between the two sets of figures is, indeed, more 

 complete than I anticipated, and hence I consider not only 

 that Mr. Sutton's conclusion to the effect that " there is 

 nothing to justify the statement that South Africa has 

 been under the same influence as that which set up the 

 prolonged drought in Australia and the dry years in India " 

 is neither in accordance with what I hold to be the general 

 meteorological conditions and relations of the whole Indo- 

 oceanic area nor even with the data which Mr. Sutton 

 furnishes. The probability, so far as I can judge, is at 

 least twenty to one that there is some relation such as I 

 have suggested. The chief object of my address was, I 

 may add, to urge the necessity for the coordination and inter- 

 comparison of the meteorological observations of the whole 

 Indo-oceanic area and their discussion as a whole by an 

 efficient scientific staff in London. The question at issue 

 between Mr. Sutton and myself, for example, could be 

 authoritatively settled by such an investigating office. 



In conclusion, I hope that my remarks may not be inter- 

 preted as in any way depreciating the value of Mr. Sutton's 

 work in collecting and discussing as a whole the rainfall 

 data of South Africa, and in utilising the data to obtain 

 normal means for purposes of comparison. His work will, 

 I am confident, be appreciated by all interested in .African 

 meteorology from any point of view. John Eliot. 



Bon Porto, Cavalaire, Var, France. 



