86 
NATURE 
[| NoveMBER 26, 1896 
Beginning with a few empirical sequences first suggested by 
Blanford, the principles on which the seasonal forecasts, issued 
semi-annually from the Simla Office, are now based, are every 
year becoming more rational and trustworthy, and recognising the 
immense practical service of such a system to the country, the 
Government have recently sanctioned the establishment of fresh 
observing stations in Persia, a cable to the Seychelles, and the 
continuation of the monsoonal charts of the Indian Ocean, which 
have already thrown considerable fresh light upon the conditions 
which regulate the normal and abnormal development of the 
weather characteristics of each monsoon in different years. 
The exceptional feature possessed by India, in common with 
most tropical countries, and which renders it possible to deal 
with its weather in broad masses and for long periods, instead of 
by the hour or day, is the fact that the periodic or climatic 
changes are far more important than those irregular, aperiodic, 
and ephemeral fluctuations which in these latitudes constitute the 
dominant components of weather. 
Thus,'to quote a single example, when all the extreme diurnal 
changes of pressure at principal observatories in India are tabu- 
lated for a year in groups according to magnitude, it is found 
that 95 per cent. of 
the regular hourly oscillation ; and the same is true of similar 
changes in the other elements. 
In brief, while daily weather anomalies and their prediction 
are the chief problem of the European forecaster, the relative 
unimportance of these in India, compared with the broad seasonal 
changes, or the variations of the average weather of a season 
from the normal (where each day resembles its predecessor so 
much that the effects of a certain type become cumulative, instead 
them are less than those which are due to | 
of being compensated by alternation), naturally leads the Indian | 
forecaster to regard the elucidation of long-period variations, 
and their prediction, as the problem which should demand his 
principal attention. 
Daily storm warnings may be of service to the shipping in 
the Indian ports, and round the coasts ; but for everything relat- 
ing to agriculture and internal land economy, seasonal forecasts 
are undeniably superior. 
With a good telegraphic system, daily forecasting can proceed 
fairly successfully on purely empirical lines, not unlike those by 
which a railway signaller is able to announce the arrival of an 
approaching train. On the other hand, successful prediction of 
circumstances not already in progress, and dependent upon 
conditions related to the movements of large air currents over 
extensive areas, and occupying considerable periods of time—if 
at first provisionally approached by empirical sequences—will 
ultimately necessitate as much rational inquiry, explanation and 
deduction as science alone can supply. India thus demands, 
and, fortunately, has hitherto succeeded in securing, a service of 
forecasters possessed of higher scientific training, as well as 
practical skill, than probably any weather service in the world. 
The preparation of the monsoon forecasts in India is based at 
present on three or four broad sequences, whose rational causes 
are as yet only partially understood. Coupled with these are a 
considerable amount of deduction from rational hypothesis, 
comparison with present and past conditions over surrounding 
areas, analogy with like conditions in previous years, and modifi- 
cation according to ascertained persistence of local anomalies. 
The determining factors may be classed as (1) local, (2) 
general. The local factors, which were formerly considered the 
most important and, indeed, the dominant causes of the mon- 
soon, have, of late years, and especially since the publication of 
Indian Ocean monsoon maps, been shown to be subordinate 
to, and of merely secondary importance compared with, those 
which evidently control the strength and quality of the monsoon 
current itself, 
For the summer, or south-west monsoon, these local factors 
are :— 
(a) The snowfall of the preceding cold weather, and (4) the 
local anomalies over India and adjacent seas during the ante- 
monsoon months shown in the monthly average anomaly maps. 
At one time the former of these was thought to be the key to 
the problem of drought or heavy rains over the whole country 
during the ensuing months, scanty rains being the sequel to heavy 
snow. Experience, however, has shown that, while heavy snow- 
fall over the Himalaya, especially late in the season, asin April 
and May, exercises an important influence in delaying the 
arrival, and checking the advance. of the monsoon over those 
areas of Upper India which border on the hills where such ex- 
cessive falls have occurred, the converse is not so effective, while 
NO. 1413, VOL. 55] 
in any case the effects are liable to be counteracted by the intrw- 
sion over India of a monsoon current of more than ordinary 
strength. This latter is a circumstance which is regulated by 
influences connected with the circulatory system of the whole 
depth of atmosphere over an entire hemisphere, and, therefore, 
quite too large to be seriously modified by the local reactionary 
effects of a comparatively small snow-covered area. 
Nevertheless, the reports of the winter snow-fall over the 
Himalaya are of considerable importance in estimating the com- 
plex probability of an early or late, favourable or unfavourable 
| monsoon, and form one of the recognised official principles on. 
which the final forecast is based in May. 
The second factor (4), once thought to be all-powerful, and still 
of considerable importance in determining the average local and. 
provincial variations of rainfall and weather during the pre- 
valence of the monsoon, is estimated by the synoptic present- 
ment of the temperature, and particularly the pressure anomalies 
exhibited in the monthly anomaly maps during the ante-monsoon, 
or *‘ hot-weather” period. 
These anomalies are found to persist more or less throughout 
the entire period of the south-west monsoon, and indicate the lines 
or zones which are favoured or avoided by the cyclonic vortices 
which distribute ‘the monsoon rains. As Mr. Blanford: 
pointed out, in his classical memoir on the rainfall of India, they 
‘rather indicate a dependence of the storm track on a quasi- 
persistent distribution of pressure than merely a modification of, 
the average pressure distribution by the passage of the barometric 
depression accompanying the storm.” 
In fact they may be compared 'to.the moulds into which molten 
metal is run, and which fashion its shape by guiding it into pre- 
existing channels and cavities. 
The circumstances which determine the sudden bursting of 
the Indian monsoon in the first week of June, have recently 
been graphically described by Mr. Eliot in his paper on the 
character of the air movement on the Indian seas and equatorial. 
belt during the south-west monsoon period, in the Quarterly 
Journal of the Royal Meteorological Society in. January of the 
present year. 
From a careful study ef the Indian monsoon charts, and the: 
barometric conditions over the equatorial region, it has been 
found that the northward! advance of the sun, and the estab- 
lishment of a thermal focus over the Indian land area, tends 
to weaken the northern side of the equatorial atmospheric crater 
which surrounds the ascensional terminus of the normal north: 
and south-east trade-wind systems, and finally stop the ascen- 
sional outlet for the latter. Its consequent horizontal outflow: 
northwards, like a pent-up lava stream,. towards the newly- 
formed ascensional focus over the Indian land area, occurs dur-- 
ing the last fortnightof May, in conjunction witha simultaneously 
sudden rise of pressure over the equator, and a consequent in- 
crease in the northward gradient. 
This sudden transformation is, moreover, effected apparently 
more by actions tending to increase the high pressure south of 
the equator, due to the seasonal enlargement of the permanent 
south polar cyclone, than by any local actions over India or 
Southern Asia, though it is possible that the general conditions. 
over the latter assist. 
It is, therefore, to the South Indian. oceanic area that the 
attention of the Indian forecasters of the monsoon is new directed, 
in order that they may have early intimation of the variations im: 
the ‘‘zzs @ tergo.” 
As soon as the current reaches the Indian land area, it fills up: 
the local inequalities in the pressure mould; and since by the 
principle of ‘* courtant ascendant” persistence, such local charac- 
teristics. once initiated, tend to continue, the forecaster is able to: 
deduce, on the supposition of a normal, excessive, or weal. 
monsoon (a distinction which can ordinarily be determined on. 
its first appearance), the probable local deviations which form. 
such a valuable practical part of the forecast. The results. 
of these two local factors, together with those of a more 
subordinate character, are all liable to serious modification. 
through the dominant influence of 
(2) The general character and strength of the monsoon: 
current. 
It is now one of the accepted ‘canons of Indian meteorology, 
which may be considered as due to the industry and perspicacity 
of its present high priest, that while the antecedent local, 
anomalies over the Indian area, introduced by thermal con- 
ditions, modify the character of the current and control its. 
local effects, they are in no sense, as was formerly believed.,, 
