It is also shown that this poison is produced by rivers, by all flat rivers at least, or those of which the progress is slow and through meadow lands; while this is pointed out as one of the causes, especially, which is not suspected or not believed in England. And here we can add a fact to our author’s statement, which is decisive: this is the case of the barracks at Morne Bruce, in Dominica, situated on a steep and rocky hill, perfectly dry, and free from all other causes of suspicion, while eternally subject to the most severe fevers. And the cause is, a mountain stream, about 300 yards below this building, in the valley, always covered by a mist in the evenings, and ascertained, by direct experience, to be the very cause of the diseases in question. Our author also notices canals, mill-ponds, ornamental waters, and all other pools and ponds, even to so small a dimension as those formed in gravel-pits; pointing out those, in particular, as common causes of fever about London, and apparently much inclined to pass a very severe judgment on the canal in St. James’s Park, and also on the pond in St. James’s Square, while apparently restrained by his prudential reasons, which appear to us sufficiently misplaced, or, as we should fairly call them, somewhat absurd. But as we must not affront a writer whose papers we have admitted, we shall say no more on this matter. In noticing drains, he also speaks of moats and modern fortifications; attempting to show that the fevers so common in the sieges of ancient castles were produced by their moats, and noticing the familiar fact of the frequency of fevers in fortified towns. Lakes also are pointed out as situations generating this poison: and it is here especially noticed that if, in those and other cases, malaria is produced by the vegetable growth and decomposition, [p105] so is it the consequence of the exposure of the mud of such receptacles of water; a cause which is again treated of at greater length in the subsequent chapter.