August 25, 1904] 



NA TURE 



403 



Variation of Mean Actual Rainfall of Period from Normal. 



1891 

 1S92 

 1S93 

 1894 



189s 

 1896 

 1897 



I goo 

 1901 

 1902 



Normal roughly. 



I inch S inches 35 inches j 41 inches 



(2) The following gives the chief features of the rainfall 

 i>f the first period, 1S92-4 : — 



(a) The excess was almost as marked in the dry as m 

 ihe w-et season. This is strongly shown in the year 1893 

 of maximum excess. 



(6) The excess w'as on the whole more strongly exhibited 

 in the field of the Bombay than of the Bengal current. 



(<-) The rainfall of the dry season was as markedly in 

 excess in Persia, Baluchistan, Afghanistan, and the 

 Himalayan area as in Northern India. 



(d) The maximum height of the Nile floods (in September) 

 was above the average. They were abnormally high in 

 1892 and 1894. 



(c) The rains were favourable over Australia and South 

 Africa during this period, according to the reports received 

 in India. 



(/) Hence, as a general inference, the rainfall was in 

 general excess in each year of the period over the Indo- 

 oceanic region, and not only in the south-west but also in 

 the north-east monsoon in Southern Asia. 



(3) The chief features of the rainfall of the second period, 

 1895-1002, in the Indo-oceanic region were as follows : — 



(a) The rainfall was as deficient relatively to the normal 

 in Ihe cold weather as in the rains or wet season. 



(()) The cold-weather or winter precipitation was almost 

 continuously in marked defect in Asiatic Turkey, Persia, 

 Afghanistan, Baluchistan, the Himalayan area, and South 

 Tibet. The opposite variation obtained in Central Asia, 

 as is shown by available data for Tashkend, Samarcand, 

 Irkutsk, and other stations. 



It) The storms of the cold weather were fewer in number 

 and feebler in character in each year of the period than on 

 the average of the preceding sixteen years 1S76-91. 



(rf) The south-west monsoon rainfall was most largely 

 in defect in the interior districts served by the Bombay 

 current. 



(e) There was a marked tendency in each year for late 

 commencement and early withdrawal of the monsoon 

 currents, and for deficient rainfall throughout the whole 

 season over the greater part of India. These features were 

 very pronounced in the years i8q6, 1800. and iqoi. 



(/) The most remarkable feature of the period was that 

 the region to the south of the equator, including South and 

 Fast Africa, Mauritius, and Australia, was similarly 

 affected. 



In India the years 1896 and 1899 were years of severe 

 drought, followed by famine over very large areas. The 

 area in which the crops failed more or less completely was 

 about 250,000 square miles in extent in 1896 and 500,000 

 square miles in iSgq. In the 1899-1900 famine upwards of 

 6.500,000 people were on famine relief for several months. 

 The loss of cattle due to failure of water and fodder was 

 very great, numbering many millions. In some districts 

 from 90 to 95 per cent, of the cattle died off from slow 



NO. 18 I 7, VOL. 70] 



starvation and want of water. In New South Wales and 

 Queensland almost continuous drought prevailed from 1896 

 to 1902. It is estimated that more than fifty millions 

 of sheep, value 12,500,000/., were lost in New South Wales 

 during these seven years of drought. 



.Mr. Hutchins, Conservator of Forests, Cape Town, states 

 that drought prevailed more or less persistently over the 

 Karoo region in South .Vfrica from 1896 to 1903, and that 

 cattle and sheep perished by millions. He also states that 

 the drought extended to British Central .'\frica from 1898 

 to 1903. 



The previous statements evidence the continuity, ex- 

 tension, and intensity of the drought. 



The Nile floods followed very closely the variations of 

 the rainfall in Western India, the floods of the years 1899 

 and 1901 were both amongst the lowest on record. This 

 shows that the rainfall in the Abyssinian region was more 

 or less generally in defect during the period and most 

 largely in the years 1899 and 1901, when the rainfall of the 

 Bombav current was very deficient. 



Hence, as a general inference, the period 1895-1902 was 

 characterised bv more or less persistent deficiency of rain- 

 fall over practically the whole Indo-oceanic area (incltjd- 

 ing Abyssinia). The economic results in the dry interior 

 districts of India, South Africa, and Australia were the 

 same— large loss of cattle and great loss of capital. The 

 drought in Southern Asia was as marked in the north-east 

 as in the south-west monsoon, and hence the variation was 

 not seasonal but general. . i j ■ 



The variations of temperature, humidity, and cloud in 

 India during the whole period were large and in direct 

 accordance with the rainfall. In other words, during the 

 period 1892-94 the air was damper with lower temperature 

 than usual, "and cloud above the normal. On the other 

 hand, from 1895 to 1902 temperature was steadily in excess, 

 cloud less than usual, and humidity below the normal. 



The most remarkable variation was that of the solar 

 radiation as indicated by observations of the solar radiation 

 thermometer (black bulb in vacuo). 



The most interesting feature of the meteorology ot tne 

 period 1892-1902 is that the variations of the solar insolation 

 are the inverse of those which might have been expected 

 from the cloud and humidity data. In other words, solar 

 radiation was in excess in the period of increased humidity 

 and cloud, and in defect during the greater part of the 

 period of drought, decreased humidity, and cloud. the 

 series of eight curves exhibited, out of a larger number- 

 prepared from the data of a number of stations in India at 

 which these observations are carefully recorded, show the 

 most important facts, and indicate that there was a con- 

 tinuous decrease of insolation on the average of all stations 

 from 1891 to 1902. The curves for Aden, Calcutta, and 

 Leh it will be seen, agree in their most important features. 

 The' observations are quite concordant and probably repre- 

 sent a most important feature of the period. They indicate 

 either a continuous and considerable decrease of emission 

 of solar energy during the period, or unusually lar-ge absorp- 

 tion in the upper atmosphere. In order to decide this 

 question comparison is necessary with similar data for other 

 lar^^e areas as, for example, Europe and North America. 

 It ts, however, clear that in India the insolation data of 

 this unique period are of exceptional interest and value. 



The preceding statements have shown that variations of 

 rainfall for prolonged periods similar in character have 

 occurred, and may hence occur again, over the very large 

 area including the Southern Asian peninsulas, East and 

 South Africa, Australia, and, perhaps, the Indian Ocean. 

 The abnormal actions or conditions giving rise to these 

 large and prolonged variations must hence be persistent for 

 long periods, and be effective over the whole of that 

 extensive area, and hence cannot be inferred with certainty 

 from the examination of the data of one small portion of 

 the area affected— e.g., India. The variations undoubtedly 

 accompany variations in the complete atmospheric circula- 

 tion over the Indo-oceanic area, and the effective forces or 

 actions must be such as to influence the whole movement 

 in a similar manner in the two monsoons or seasons of 

 inverse conditions in Southern .'\sia. This inference 

 furnishes a very strong reason for the conclusion that the 

 meteorology of the whole area similarly affected from i8q2 

 to iqo2 should be studied as a whole, and not in fragmentary 



