So abundant are fish in all parts of the island, that Knox says, not the running streams alone, but the reservoirs and ponds, “nay, every ditch and little plash of water but ankle deep hath fish in it."[1] But many of these reservoirs and tanks are, twice in each year, liable to be evaporated to dryness till the mud of the bottom is converted into dust, and the clay cleft by the heat into gaping apertures; yet within a very few days after the change of the monsoon, the natives are busily engaged in fishing in those very spots and in the hollows contiguous to them, although the latter are entirely unconnected with any pool or running streams.  Here they fish in the same way which Knox described nearly 200 years ago, with a funnel-shaped basket, open at bottom and top, “which,” as he says, “they jibb down, and the end sticks in the mud, which often happens upon a fish; which, when they feel beating itself against the sides, they put in their hands and take it out, and reive a ratan through their gills, and so let them drag after them."[2]

[Footnote 1:  Knox’s Historical Relation of Ceylon, Part i. ch. vii.  The occurrence of fish in the most unlooked-for situations, is one of the mysteries of other eastern countries as well as Ceylon and India.  In Persia irrigation is carried on to a great extent by means of wells sunk in line in the direction in which it is desired to lead a supply of water, and these are connected by channels, which are carefully arched over to protect them from evaporation.  These kanats, as they are called, are full of fish, although neither they nor the wells they unite have any connection with streams or lakes.]

[Footnote 2:  Knox, Historical Relation of Ceylon, Part i. ch vi.]

[Illustration:  FROM KNOX’S CEYLON, A.D. 1681]

This operation may be seen in the lowlands, traversed by the high road leading from Colombo to Kandy.  Before the change of the monsoon, the hollows on either side of the highway are covered with dust or stunted grass; but when flooded by the rains, they are immediately resorted to by the peasants with baskets, constructed precisely as Knox has stated, in which the fish are entrapped and taken out by the hand.[1]

[Footnote 1:  As anglers, the native Singhalese exhibit little expertness; but for fishing the rivers, they construct with singular ingenuity fences formed of strong stakes, protected by screens of ratan, that stretch diagonally across the current; and along these the fish are conducted into a series of enclosures from which retreat is impracticable.  MR. LAYARD, in the Magazine of Natural History for May, 1853, has given a diagram of one of these fish “corrals,” as they are called, of which a copy is shown on the next page.]

So singular a phenomenon as the sudden re-appearance of full-grown fishes in places that a few days before had been encrusted with hardened clay, has not failed to attract attention; but the European residents have been content to explain it by hazarding conjectures, either that the spawn must have lain imbedded in the dried earth till released by the rains, or that the fish, so unexpectedly discovered, fall from the clouds during the deluge of the monsoon.