19 DOVE ON THE LAW OF STORMS. 
the 15th of May and the 15th of June*. Throughout the greater 
part of its course “it was a gale or strong wind blowing with 
tolerable steadiness from one quarter of the compass,” and it 
was only at a particular part that it was a hurricane or violent 
wind blowing in a circle or vortex of greater or less diameter+. 
It blew as a violent south-west monsoon in the space between 
the east coast of Ceylon and Masulipatam, and across the bay 
of Bengal towards the mountain range of Arracan, where it 
turned completely at right angles, advancing into the interior, 
up the Ganges, and blowing as a south-eastern current over 
Calcutta and Benares to Cawnpoor, Lucknow and Agra. At 
the place of flexure near Arracan,—in the focus, as Piddington 
expresses it, of the parabolic course of the storm,—a whirlwind 
originated and advanced parallel to the coast, and passing off 
the mouths of the Ganges, moved in a direction from between 
E.N.E. and E. towards W.S.W. and W., from Shapooree Island 
towards Vizagapatam, Gangam, Juggernaut, and the mouths of 
the Mahanuddy and Bramnee, rotating like the West India 
hurricanes in the sense 8. E. N. W. 
Here we have a whirlwind rotating in a precisely similar man- 
ner to those before described, arising (under circumstances ori- 
ginally quite different) when the direction of the storm on its 
eastern side was more towards the north than on its western 
side; and possibly the typhoons of the Chinese sea may owe their 
origin to similar causes. In the complex phenomenon of the 
south-west monsoon the conditions are analogous. The wind, 
from being south-west in the Indian sea and the bay of Bengal, 
becomes more nearly south in the Chinese sea; more extended 
observations are required to show whether this deflection is 
caused by the chain of the Philippines, or whether it is an im- 
* According to the observations of Brown it commenced at Anjarakandy on 
the Malabar coast in 1820-1833, on May 20, 31, 31, 27, June 15, May 21, 
June 18, May 26, June 5, May 9, 26, June 16, 2, 6: at Canton it set in, 
according to the Canton Register, from the 20th to 28th of April in 1830, from 
the 7th to 17th of April in 1831, from the 4th to 7th of April in 1832, from 
the 9th to 14th of April in 1833, from the 3rd of April to 8th of May in 1834, 
and from the 8th to the 21st of April in 1835. 
+ This seems to be also the case with the storms which accompany the change 
from the south-west to the north-east monsoon. These storms, which the 
Spaniards in Manilla call ‘los temporales,” are not accompanied by rain, 
but the air is everywhere darkened by the salt spray from the sea. On the 
coast of Coromandel these storms are termed “the breaking out of the mon- 
soon.” On the Malabar coast the Portuguese call those which are peculiarly 
violent, ‘* Elephanta.” 
