266 



KNOWLEDGE. 



[Deoembee 1, 1896. 



amount really means, let us suppose the excess rainfall of 

 1893 to be collected in a tank with a scjuaro base of eight 

 miles a side. Then if its altitude were that of the snow 

 line of the Himalaya, viz., seventeen thousand feet, such 

 a tank would barely contain the total volume of excess 

 water which fell over India during that year. Similarly, if 

 in order to supply the defect in 1868 we imagine a hose 

 pipe to stretch all the way from the earth to the moon, of 

 half an acre in section and full of water, it would represent 

 a trifle more than the above defect. If it were required 

 to irrigate the country with this hose, in order to keep up 

 the supply to the normal during the six months of the 

 monsoon, the water would have to issue from the hose at 

 the rate of fifty-five miles per hour continuously. 



Moreover, although such variations, spread over the 

 entire area, reach this gigantic amount, their local effect is 

 relatively much greater, and produces much more disastrous 

 effects in the dry zone inland than near the coast, where 

 the rainfall is normally high. In this zone, which com- 

 prises the Deccan, Mysore, South Madras, Central and 

 South Punjab, and the western section of the North- 

 West Provinces, variations often occur amounting to 

 several hundred per cent, of the normal supply. It is this 

 district which, as at present, is most liable to famine- 

 producing droughts or floods. 



The method of long period prediction began some years 

 back imder the late Mr. lUanford, F.R.S., by the recog- 

 nition of certain sequences which were observed to occur 

 in the summer monsoon rains, according as the snowfall 

 of the preceding winter on the Himalayan slopes was 

 heavy or otherwise. A heavy winter snowfall was usually 

 found to be followed by a light summer monsoon rainfall, 

 and rice versih 



Although this factor has latterly been found to be liable 

 to some uncertainty and modification, owing to variations 

 in the absolute strength, quality, and duration of the 

 monsoon current itself, it still forms one of the leading 

 principles by which the extension of the current to its full 

 northerly limits is predicted. 



Heavy and untimely snowfall, especially in April or 

 May, exercises a powerful influence in preventing or 

 delaying the extension of the monsoon over Upper India. 



Another factor is afforded by the local peculiarities 

 which are manifest in the hot weather period immediately 

 preceding the arrival of the monsoon, and which arc best 

 estimated by means of the synoptic variations of the 

 current barometric pressure from the normal. 



At one time it was thought that these local anomalies 

 of pressure were the chief cause of the monsoon itself, on 

 the principle of the sea breeze towards a heated area. 

 This was formerly alluded to as the " furnace theory." 

 It is now, however, clearly recognized that the advance of 

 the massive current itself is not directly the result of the 

 hot weather over the Indian area, but that it forms part 

 of a larger system of circulation. Its forward extension 

 across the Equator is as much due to a vis a teiyo, effected 

 in the South Indian Ocean, as to a rix ad frontem, in con- 

 sequence of changes over the Indian land area following 

 the northward march of the sun. As soon, however, as 

 the current is once established over India, it is found that 

 it tends to concentrate its rainfall energy mostly over those 

 districts where the pressure was lowest during the ante- 

 monsoon months. 



Hence this second factor is of great service in deter- 

 mining the local or provincial variations of rainfall, 

 though, like the snowfall factor, the deductions have to 

 be modified in correspondence with the precise character 

 of the incoming current. 



The last factor is the condition of the south-east trade 



wind of the South Indian Ocean, which is now found to be 

 nothing else than the direct parent of the monsoon. 



The tuiidus opernnili by which the southern trade 

 wind crosses the Equator in May and rushes across 

 the Indian Ocean as the south-west monsoon, which 

 breaks on the Indian coasts early in June, has been 

 graphically described by Mr. Eliot in a recent paper in 

 the Quarterbi .hnirnal af the Boi/nl Meteanilogicid Society 

 (January, 1890). We need not here allude to it further 

 than to say that while a feeble sea breeze develops along 

 the Indian coasts for some time previous to the monsoon 

 burst, the latter is evidently in no sense its culminating 

 stage. On the contrary, as Mr. Eliot substantially puts 

 it, it is plainly due to the breakdown of the upilow 

 over the equatorial calm belt, which allows the south-east 

 trade wind to continue its horizontal flow across the 

 Equator. It thus brings the vapour accumulated during 

 its southern journey in a massive sheet, which condenses 

 when carried up the Indian ranges and plateaux. 



The only direct means of estimating the probable 

 strength and character of this vapour current is to 

 ascertain the character of the south-east trade wind of the 

 Indian Ocean, south of the Equator, during April and May. 

 It is found that this is usually maintained comparatively 

 unaltered for some months, and is transmitted, pari ijradu, 

 to its defluent extension, known as the south-west 

 monsoon of India. 



Up to the present time the information has usually 

 been derived either from the logs of ships arriving at the 

 Indian ports from the southern seas or else by letter from 

 Mauritius, the Seychelles, etc. Latterly, however, the 

 Indian Government has sanctioned the establishment of 

 cable communication to the Seychelles, and is preparing 

 to do everything in its power to accelerate the trans- 

 mission of news from the Mauritius, Zanzibar, etc. 



Meanwhile, as a supplement to the direct observation of 

 the trade current, it has been found that the monthly 

 average barometric pressure over India is subject to 

 periodic, long oscillations, above and below the mean. 

 These vary from six to twenty-four months, and are usually 

 some multiple of six months. 



These waves of pressure are found to be connected with 

 the development of the monsoon current in such a way 

 that if the wave is rising during the month or two (April 

 and May) preceding the south-west monsoon, the rainfall 

 will be scanty, and the reverse if it is falling. On the 

 other hand, if it is just beginning to rise during the month 

 preceding the winter monsoon (November), the rains 

 which fall in December and January will probably be above 

 the average, and vice rersd. These waves occur, reversed 

 in phase, oa the south side of the Equator, and indicate, 

 as Mr. Eliot (the head of the Indian Weather Bureau) 

 says, checks and accelerations in the seasonal mass 

 transfer of air across the Equator between the Indian 

 Ocean and Southern Asia. 



Twelve such waves have occurred during the past twenty 

 years, and their careful study seems destined to open out 

 a new departure in meteorology by permitting seasonal 

 forecasts to be projected on a rational basis. 



Regarding the forecast this year, though no direct 

 reference was made to the possibility of a famine-producing 

 drought, attention was drawn to the signs of a weak mon- 

 soon from the observations, particularly at the Seychelles. 



Up to the present time the probability of a break in the 

 middle of the rains or their early termination in any year 

 are admitted to be difficult to determine with accuracy. 

 The latter is, perhaps, one of the most important relations 

 to be able to forecast, since the present disastrous scarcity 

 is directly traceable to an unusual scorching in September, 



