throughout the Jheels, where fevers and agues are rare; and though no
situations can appear more malarious to the common observer than
Silhet and Cachar, they are in fact eminently salubrious. These facts
admit of no explanation in the present state of our knowledge of
endemic diseases. Much may be attributed to the great amount and
purity of the water, the equability of the climate, the absence of
forests and of sudden changes from wet to dry; but such facts afford
no satisfactory explanation. The water, as I have above said, is of a
rich chesnut-brown in the narrow creeks of the Jheels, and is golden
yellow by transmitted light, owing no doubt, as in bog water and that
of dunghills, to a vegetable extractive and probably the presence of
carburetted hydrogen. Humboldt mentions this dark-coloured water as
prevailing in some of the swamps of the Cassiquares, at the junction
of the Orinoco and Amazon, and gives much curious information on its
accompanying features of animal and vegetable life.

The rains generally commence in May: they were unusually late this
year, though the almost daily gales and thunder-storms we
experienced, foretold their speedy arrival. From May till October
they are unremitting, and the country is under water, the Soormah
rising about fifty feet. North-easterly winds prevail, but they are a
local current reflected from the Khasia, against which the southerly
perennial trade-wind impinges. Westerly winds are very rare, but the
dry north-west blasts of India have been known to traverse the delta
and reach this meridian, in one or two short hot dry puffs during
March and April. Hoarfrost is unknown.* [It however forms further
south, at the very mouth of the Megna, and is the effect of intense
radiation when the thermometer in the shade falls to 45 degrees.]

China roses and tropical plants (_Bignoniae, Asclepiadeae,_ and