NoveMBER 26, 1896] 
WN AMA Ee Fe 
87 
‘its proximate cause, and that the changes in its general strength 
and character from year to year are more the result of actions 
taking place south of the equator than of any peculiar conditions 
over the south Asiatic land area. 
The extension of the area of observation, as far as Mauritius 
and the Seychelles, is the logical outcome of this principle ; and 
though the information at present obtainable is mostly empirical 
in form, it is found that the essential subsequent unity of the 
south-east trade and the south-west monsoon enables early 
information of the character of the former in the southern seas 
to be used as an empirical index to the seasonal character of the 
latter as soon as continuity has been established across the 
equator. 
A strong trade-wind argues far? fassu a strong monsoon, and 
therefore a good rainy season over India, except where it is 
counteracted by opposing local factors. 
It will probably be some time before rational knowledge of 
this important factor will supersede the empirical for the 
purposes of practical forecasting. Meanwhile every extension 
of the means of accelerating the transmission of news of south- 
east trade conditions is invaluable in forming a trustworthy 
seasonal forecast from the Simla Office. 
The summer monsoon forecast is made up provisionally about 
the fourth week in May, and held over until symptoms of the 
coming monsoon manifest themselves in Bombay, in order to 
allow of the latest information being incorporated. June 6 or 
7 is the average date for Bombay, and it is frequently from two 
to three weeks later before it reaches the Punjab. 
The work of preparation is no light matter, since the greatest 
care is taken by Mr. Eliot to allow for every factor, and to arrive 
at a fair balance of probabilities. 
It is a serious matter to forecast for six months over an area 
half aslarge asthe United States, and the work occupies a week 
after every map and datum is on the table. 
Two points on which the agricultural value of the monsoon 
rainfall largely depends are at present only partially predictable, 
viz: (1) the probability of a prolonged break in the rains in 
July or August, and (2) the probability of an unusually early 
termination of the rains in Upper India or Bengal. 
The former depends chiefly on the relative strength of the 
two branches of the monsoon current, the break generally 
occurring with a relatively weak Arabian Sea current, while the 
latter depends on the early establishment of the high pressure 
over North-west India and North Burmah, which causes the 
reversal of the gradient, and, as it were, drives the monsoon 
out of the Bay of Bengal. 
These conditions can only be inferred months before their 
occurrence by analogy with previous years presenting similar 
characteristics. 
Once they have started, however, they can be employed in 
determining the probability of the early or late occurrence of 
the rainfall of the winter monsoon. 
The rainfall of this monsoon, though very inconsiderable com- 
pared with that of the summer, is, agriculturally speaking, of 
‘great value, since upon its presence the entire fortunes of the 
rabi crop depend. 
Originating, as Mr. Eliot has so exhaustively shown in his 
recent ‘‘ Memoir on the Winter Storms of India,” in a lofty 
current 10,000 feet above sea-level, having no lower oceanic 
continuation, and shedding its vapour in storms bred on the 
plateaus of Afghanistan and Persia, it is at present impossible to 
obtain direct information of the character of the winter monsoon 
before its descent to the North Indian plains in December. A 
divining-rod has, however, been lately discovered, by which it 
may be inferred from the nature of the vertical pressure anomalies 
for the months immediately preceding December. 
These anomalies represent the departure from the mean of the 
monthly mean pressure at stations near the foot of the hills, 
minus the corresponding departures at the mountain stations 
7000 feet above. When the differences (plain minus hill stations) 
are positive, the inference is that the subsequent winter will be a 
dry and stormless one ; if negative, precisely the reverse. 
Since the character of the winter monsoon is found to be re- 
markably constant all through, probably because being an upper 
current it is unaffected by local and land influences, the forecast 
from this empirical sequence alone is found to be remark- 
ably trustworthy. 
There are some additional sequences, first indicated some years 
ago, which are useful in confirming the conclusions derived from 
the vertical anomalies, such as the probability of a light winter 
NO. 1413, VOL. 55 | 
fall succeeding a light and early departing monsoon, and véce 
versd. Such sequences are, however, really included under a 
more general law which, though at present empirical in form, 
appears likely to lead to a rational explanation of the chief yearly 
variations in both monsoons. This law is the outcome of a 
recent discussion, by Mr. Eliot, of certain oscillatory variations of 
pressure which are found to be common to the entire Indian 
oceanic and continental areas. 
It has been found that the monthly mean barometric pressure 
of India is subject to a series of long-period waves of nearly 
equal amplitude, ranging up to as much as ‘07 inch (which is 
up to drought-producing power in India), and varying from 
twelve to twenty-four months in length. Twelve of these 
occurred in the last twenty years. 
They are found to occur completely reversed in phase at 
Mauritius, and are considered to represent the major fluctuations 
in the annual oscillatory flow of air to and from the South Indian 
Ocean and India, in the form of the monsoons, together with 
similar conditions involved in the corresponding return flow in 
the upper and lower atmosphere, according as it is summer or 
winter. They are likewise precisely opposite in phase to the 
vertical anomalies. In other words, these oscillations of pressure 
represent compensatory variations in the horizontal transfer of 
air across the equator, and in the vertical transfers at the 
northern and southern termini of the circulatory system, which 
are intimately bound up with the strength and character of the 
monsoons. 
The principal maxima pressures usually occur about the 
vernal, and the principal minima about the autumnal equinox. 
Employed as an empirical sequence for forecasting, the rule may 
be stated thus. 
In years when the sea-level pressure anomaly is such that the 
curve is a descending slope during the spring months, the con- 
clusion is that the south-west monsoon will be one ofexcessiverain- 
fall, and vce vers? when the slope is ascending, it will be com- 
paratively dry. Conversely, the years of heavy winter rainfall 
during the north-east monsoon, tend to coincide with the maximum 
epochs of these pressure anomaly waves, z.¢. with the minimum 
epochs of their vertical anomaly analogues, probably because 
diminished pressure above coincides with increased pressure 
below. V2ve versd, years of light rainfall coincide with the epochs 
of general minimum anomaly at sea-level. By the accumula- 
tion of such sequences, and their gradual determination in a 
rational form, the science of long-period forecasting is being 
built up in India 
The occurrence of a severe scarcity during the present year 
is a timely commentary on the practical value and limits of the 
monsoon forecast. 
Of the four causes detailed by Mr. Eliot in his paper on 
**Droughts and Famines in India,” read before the Meteoro- 
logical Congress at Chicago in 1893, the present famine is due 
to the last—viz. ‘‘ Unusually early termination of the south-west 
monsoon rains. This is especially fatal in the case of rice crops 
on unirrigated land ; 
The same circumstance is also peculiarly prejudicial to the 
sowing of the winter or “‘rabi” grain crop which is reaped in 
March, especially when it is succeeded by scanty winter rains 
in December and January. , 
As has been already observed, this early termination of the 
south-west monsoon is one of the conditions which, at present, 
lies outside the conventional forecast, though it is rapidly be- 
coming manifest that it is dependent on the general state of the 
south-east trade wind of the Indian Ocean. In the Foregast 
Circular issued this year by Mr. Eliot, and dated Simla, June 3, 
attention is drawn fo the fact that, during the past two or three 
years, the “causes of the large variations of the rainfall in 
India have been evidently due to abnormal conditions outside 
the Indian area, and not to local peculiarities or abnormal 
meteorological features in India itself.” 
The anomalies are so remarkable that they are worth re- 
producing, as in the following table. 
Rainfall anomaly from 
Percentage varia- 
tion from 
Years. normal over India 
in inches. mean. 
1893 +8:94 ... a +22 
1894 +6748... ee +16 
1895 — 2°90 : Sh 
The pre-monsoon conditions of temperature, pressure, and 
snowfall were almost identical in 1894 and 1895. 
The south-east trade wind, however, was weak in 1895, and 
