constructed as shown in Figs. 3 and 15, were invented in Germany in the latter part of the sixteenth century; the exact date, as well as the inventor's name, is uncertain. Bellows working on the principle of those used in accordions and concertinas have also been known for many centuries. An engraving, showing such bellows in use blowing a furnace, is given in the great work of Agricola,[1] who also illustated rotary fan-blowers; but these evidently did not propel the air centrifugally, as does the modern fan-blower, but pushed the air forward, very much as a revolving paddle-wheel pushes water.
Another very curious apparatus for blowing furnaces and smiths' fires is called a trompe. It consists of a vertical pipe, usually made of wood, of a length suited to the fall of water. Near the top of this pipe there are pierced a number of comparatively small lateral openings which incline downward in their passage through the thickness of the sides of the pipe, whose lower end enters the closed top of a barrel or other air-tight vessel, from which proceeds a tube to convey the air to the furnace or forge. This contrivance operates as follows: The descending column of water in the pipe draws in air through the lateral openings near its top, and this air is carried down by the water and separates from it in the interior of the barrel and then passes to the forge by the discharge-pipe, the water escaping through a hole at or near the bottom of the barrel. Percy,[2] speaking of this very simple blowing apparatus, says, "It is said that it was invented in Italy in 1640." But it must have originated at a much earlier date, as Branca[3] gives three applications of it, illustrated by engravings, and it is very probable that this highly ingenious method of employing the fall of water to compress air was known and used hundreds of years before the time of Branca.
The early American forges and furnaces were blown either by the ordinary leather bellows (Fig. 1), or by wooden cylinders called "blowing-tubs," or by the trompe just described, and there are still to be found in use a few examples of each of these primitive methods of "raising the wind." In Fig. 2 we have an illustration of a pair of "blowing-tubs" such as Overman[4] describes as "the best form of wooden blast-machine." The figure shows a vertical section through the axes of the upright "blowingtubs," a a, and the "wind-chest," b, placed immediately above them. Air enters the tubs from beneath and is purnped by the pistons d d, with the aid of the "clack-valves" shown in the figure, into the windchest. The pressure of the air in the windchest is determined by the weight h suspended at the lower end
