A5O REPORT—1904. 
on the rainfall resulting from the initial pressure conditions—an example of the 
persistence of meteorological conditions and actions which is a prominent 
feature of Indian meteorology. The total rainfall of each of the three areas may 
differ considerably from the normal, but there may be partial or complete compen- 
sation on the whole. Thus it is the general (but not the invariable) rule that the 
rainfall variations in Burma and Assam are usually inverse to those of North- 
Western India and also of India as a whole. 
(7) The distribution of the rainfall in any one of the three competing areas 
(but more especially in India as the largest) may vary widely from the normal— 
considerable deficiency in some areas accompanying considerable excess in others. 
This in India is undoubtedly due to local conditions—e.g., local excess or deficiency 
of pressure at the commencement of the period and established during the previous 
hot weather. These pressure variations usually accompany abnormally prolonged 
and heavy snowfall or very scanty snowfall in the Western Himalayas. 
(8) Loeal or general drought in India during the south-west monsoon may 
hence be due to— 
(a) General weakness of the south-east trades circulation. 
(6) Diversion of an unusually large proportion of the south-east trades to 
South-East or East Africa during the monsoon period. 
_ fe) Larger diversion than usual of the monsoon currents to Burma or Abys- 
sinia. 
(d) Very unequal distribution in India itself, due to local conditions esta- 
blished during the antecedent hot weather. 
These factors are given in the probable order of their importance. 
(9) Scanty rainfall or drought during the dry season or north-east monsoon in 
Northern India results from absence or unusual feebleness of the cold weather 
storms which are the sources of rainfall at that time. 
(10) The most prolonged and severe droughts in North-Western and Central 
India are due to the partial or complete failure of the rainfall of at least two 
seasons in succession. 
(11) As the two circulations in the Indian oceanic region have a common goal 
in the dry season (more especially from December to March), it is probable that 
variations in the strength of one circulation (more especially the larger) will 
modify the field and strength of the other circulation. It appears that this rela- 
tion would be shown mest strongly between the southern circulation and the 
upper movement of the northern circulation. And, as cold weather storms are 
disturbances in that upper movement, it is possible—if not probable—that the 
larger variations in the number and intensity of the cold-weather storms and 
the amount of the cold weather precipitation may be related to conditions in 
the south-east trades regions. 
(12) There appears to be little or no relation between the position and intensity 
of the Central Asian anticyclone and the number of the cold-weather storms 
and rainfall of Northern India in any season. 
The meteorology of the period 1892-1902 is of especial interest for its con- 
firmation of the above inferences, more especially the phenomena of the variations 
of rainfall in India and the causes or actions to which they are due. The year 
1891 was noteworthy for a severe local famine in Rajputana and the adjacent 
districts to the north and east consequent on prolonged and excessive snowfall in 
the Western Himalayas during the winter of 1890-91. The following gives a brief 
nay of the more prominent feature of the meteorology of this unique 
period :— 
1. The eleven-year period 1892-1902 corresponds in length to the sun-spot 
period, and it may be divided into two periods of unequal length—a short period 
of excessive rain and a long period of deficient precipitation. The maximum of 
the first period was in 1893. The second period had three strongly marked 
minima in 1896, 1899, and 1901, that of 1899 being the absolute minimum, The 
following table gives, for convenience of reference, data of the mean annual and 
