The increased attention which has been paid within recent years to fishery statistics has revealed, amongst other troublesome things, that the instruments employed play an important part both as to quantity and quality in the result. It is not necessary to discuss whether any of the instruments as now used give a fair sample of the contents of the water or ground, but it certainly is the case that different apparatus will give different samples. Let a beam-trawl and an otter-trawl work for a year over the same ground alongside drift-nets and fixed trammel-nets or gill-nets, and we can guarantee that the results will differ from one another. In the case of the drift-nets we should have a few forms probably in large quantity, in the others a great variety of forms, but in different proportions in each. And again, those obtained by the drift-nets would be practically absent from the trawls. It thus behoves the naturalist to make every kind of fishing apparatus subservient to his use, if he desires to obtain even an approximate measure of fish-life in the sea.