The number of known British fishes may be safely assumed to be less than 250, and Mr. Yarrell enumerates only 226, Dr. Cantor’s valuable work on Malayan fishes enumerates not more than 238, while Dr. Russell has figured only 200 from Coromandel.  Even the enormous area of the Chinese and Japanese seas has as yet not yielded 800 species of fishes.

The large extent of the collection alone, then, renders it of great importance; but its value is immeasurably enhanced by two circumstances,—­the first, that every drawing was made while the fish retained all that vividness of colouring which becomes lost so soon after its removal from its native element; second, that when the sketch was finished its subject was carefully labelled, preserved in spirits, and forwarded to England, so that at the present moment the original of every drawing can be subjected to anatomical examination, and compared with already named species.

Under these circumstances, I do not hesitate to say that the collection is one of the most valuable in existence, and might, if properly worked out, become a large and secure foundation for all future investigation into the ichthyology of the Indian Ocean.

It would be very hazardous to express an opinion as to the novelty or otherwise of the species and genera figured without the study of the specimens themselves, as the specific distinctions of fish are for the most part based upon character; the fin-rays, teeth, the operculum, &c., which can only be made out by close and careful examination of the object, and cannot be represented in ordinary drawings however accurate.

There are certain groups of fish, however, whose family traits are so marked as to render it almost impossible to mistake even their portraits, and hence I may venture, without fear of being far wrong, upon a few remarks as to the general features of the ichthyological fauna of Ceylon.

In our own seas rather less than a tenth of the species of fishes belong to the cod tribe.  I have not found one represented in these drawings, nor do either Russell or Cantor mention any in the surrounding seas, and the result is in general harmony with the known laws of distribution of these most useful of fishes.

On the other hand, the mackerel family, including the tunnies, the bonitos, the dories, the horse-mackerels, &c., which form not more than one sixteenth of our own fish fauna, but which are known to increase their proportion in hot climates, appear in wonderful variety of form and colour, and constitute not less than one fifth of the whole of the species of Ceylon fish.  In Russell’s catalogue they form less than one fifth, in Cantor’s less than one sixth.

Marine and other siluroid fishes, a group represented on the continent of Europe, but doubtfully, if at all, in this country, constitute one twentieth of the Ceylon fishes.  In Russell’s and Cantor’s lists they form about one thirtieth of the whole.