In the course of twenty-four successive hours twenty-seven inches of rain fell, thirty-five fell in twenty-six hours, seven inches in one hour and a-half, being the heaviest fall on record. At Poonah, on the 3d of August 1852, after a very heavy fall of rain, multitudes of fish were caught on the ground in the cantonments, full half a mile from the nearest stream. If showers of fish are to be explained on the assumption that they are carried up by squalls or violent winds, from rivers or spaces of water not far away from, where they fall, it would be nothing wonderful were they seen to descend from the air during the furious squalls which occasionally occur in June." Sir E. Tennent adds the following examples:--"I had an opportunity, on one occasion only, of witnessing the phenomenon which gives rise to this popular belief. I was driving in the cinnamon gardens near the fort of Colombo, and saw a violent but partial shower descend at no great distance before me. On coming to the spot, I found a multitude of small silvery fish from one and a half to two inches in length, leaping on the gravel of the high road, numbers of which I collected and brought away in my palankin. The spot was about half a mile from the sea, and entirely unconnected with any watercourse or pool. "Mr Whiting, who was many years resident at Trincomalee, writes me that he 'had been often told by the natives on that side of the island that it sometimes rained fishes; and on one occasion (he adds) I was taken by them, in 1849, to a field at the village of Karrancotta-tivo, near Batticaloa, which was dry when I passed over it in the morning, but had been covered in two hours by sudden rain to the depth of three inches, in which there was then a quantity of small fish. The water had no connexion with any pond or stream whatsoever.' Mr Cripps, in like manner, in speaking of Galle, says: 'I have seen in the vicinity of the fort, fish taken from rain-water that had accumulated in the hollow parts of the land that in the hot season are perfectly dry and parched. The place is accessible to no running stream or tank; and either the fish, or the spawn from which they were produced, must of necessity have fallen with the rain.'"[77] Mr J. Prinsep, the eminent secretary to the Asiatic Society of Bengal, found a fish in the pluviometer at Calcutta, in 1838.[78] It is a highly curious fact that the pools, reservoirs, and tanks in India and Ceylon are well provided with fish of various species, though the water twice every year becomes perfectly evaporated, and the mud of the bottom becomes converted into dust, or takes the condition of baked clay, gaping with wide and deep clefts, in which not the slightest sign of moisture can be detected. This is the case with temporary hollows in the soil, which have no connexion with running streams or permanent waters, from which they might be supposed to receive a fresh stock of fish.