principally by means of another ingredient, the slags of theiron furnaces; and glass of this composition is found notto be convertible into porcelain. The failure of this kindof glass serves to confirm the general results drawn fromthe former experiments; that earthy and metallic glassesmade without saline matter, are not susceptible of thischange; and that the change depends on the saline sub-stance contained in the vegetable aihes.

Page 314. Machines for blowing air by a fall of water.

I have received an account, from a worthy corre-spondent in Swifferland, of a machine which he has con-structed for a smelting furnace according to the foregoingdirections : he fays, it has so much the advantage of allother kinds of bellows, that it deserves to be introduceduniversally wherever the situation of the place will permit.The only inconvenience he finds in it is, that the cullenderand gratings are liable to be stopt up by leaves, &c. Withregard to the cullender, the obstruction may be obviatedby enlarging the holes. The gratings ought to be of alarge surface : the wire grating in the cistern on the topmay be a cylinder nearly as large as the cistern will re-ceive, for if it is no more than sufficient to cover themouth of the pipe, it will doubtless be soon choaked up:when so much of the cylinder becomes stopt, that thewater has no longer a free passage through, it may belifted up and cleaned, another being placed in the room. ofit, without the trouble of turning off the water,, or inter-rupting the going of the machine. The gratings herecan be liable to no other inconveniences, than those whichare common in other water machines, mills, aquæducts, 6ec.

Some further improvements have occurred in the con-struction of these machines, by which they may be made

effectuai-