He recommends the placing of the fish entire, if of moderate size, in a tin or plated copper dish adapted to the form and size of the fish, but a little deeper than its thickness, so as to retain all the juices, which on exposure to the heat will flow out; the surface to be lightly spread with butter with a morsel or two added, and the dish placed before the fire in a Dutch or American oven, or the special apparatus made by Burton of Oxford Street, which was exhibited at the lecture. To this I may add, that if a closed oven be used, Rumford’s device of a false bottom, shown in Fig. 3, p. 72 (see next chapter), should be adopted, which may be easily done by simply standing the above-described fish-dish, on any kind of support to raise it a little, in a larger tin tray or baking-dish, containing some water. The evaporation of the water will prevent the drying up of the fish or of its natural gravy; and if the oven ventilation is treated with the contempt I shall presently recommend, the fish, if thick, will be better cooked and more juicy than in an open-faced oven in front of the fire. This reminds me of a method of cooking fish which, in the course of my pedestrian travels in Italy, I have seen practised in the rudest of osterias, where my fellow-guests were carbonari (charcoal burners), waggoners, road-making navvies, &c. Their staple ‘_magro_,’ or fast-day material, is split and dried codfish imported from Norway, which in appearance resembles the hides that are imported to the Bermondsey tanneries. A piece is hacked out from one of these, soaked for awhile in water, and carefully rolled in a piece of paper saturated with olive oil. A hole is then made in the white embers of the charcoal fire, the paper parcel of fish inserted and carefully buried in ashes of selected temperature. It comes out wonderfully well cooked considering the nature of the raw material. Luxurious cookery _en papillote_ is conducted on the same principle and especially applied to red mullets, the paper being buttered and the sauce enveloped with the fish. In all these cases the retention of the natural juices is the primary object. I should add that Sir Henry Thompson directs, as a matter of course, that the roasted fish should be served in the dish wherein it was cooked. He suggests that ‘portions of fish, such as fillets, may be treated as well as entire fish; garnishes of all kinds, as shell-fish, &c., may be added, flavouring also with fine herbs and condiments according to taste.’ ‘Fillets of plaice or skate with a slice or two of bacon; the dish to be filled or garnished with some previously-boiled haricots,’ is wisely recommended as a savoury meal for a poor man, and one that is highly nutritious. A chemical analysis of six-pennyworth of such a combination would prove its nutritive value to be equal to fully eighteen-pennyworth of beefsteak.