consequence is, that a great fall of snow coming at the same time that the mills have not been in a state to prepare the ditches to receive the overplus water which it occasions, an inundation generally takes place in the fens; and, as the waters rise very rapidly under these circumstances after a thaw, it frequently occurs, that when the mills are set at liberty from the effects of ice, they are for some days incapable of successfully opposing the accumulation of water. On the other hand, by adopting the means of steam, the engines would be working in full effect during the continuance of a frost, and the ditches being kept proportionably low, would at all times be capable of discharging the water, and thus prevent inundation.

Mr. S. concludes this part of a very useful paper on the subject, by observing, that "as to a district of country which requires draining without any engines upon it, at the time of its being undertaken, it is a matter of doubt in my mind, whether it could not be drained more economically by steam, than by the means usually adopted, although the expence of fuel must certainly be very great. Taking the average of winds, the mills in the winter season do not throw so much water in a week, as they would in one third of the time, if they went with all the velocity of which they are capable. It follows, that one steam engine, with equal powers, would do as much execution in the

course of a season, as three windmills, and consequently a great saving would accrue in the first expense, and afterwards in attendance and repairs.'

* Vide Communications to the Board of Agriculture, vol. iv. p. 52.

course of a season, as three windmills, and consequently a great saving would accrue in the first expense, and afterwards in attendance and repairs."*

* Vide Communications to the Board of Agriculture, vol. iv. p. 52.

an atmospheric engine, belonging to the University of Glasgow, which he had undertaken to repair. In the course of his experiments with it, he found the quantity of fuel and injection water it required, much greater in proportion than in the larger engines; and it occurred to him, that this must be owing to the cylinder of this small model exposing a greater surface in proportion to its contents, than was effected by larger cylinders. This he endeavoured to remedy, by employing non-conducting substances for those parts of the engine which came in immediate contact with the steam. After a variety of experiments, the results of which we shall presently describe, he succeeded in constructing a working model, capable of producing a force equal to fourteen pounds on every inch of the piston, and which did not require more than one third of the steam used in the common atmospheric engine to produce the same effect.

It will be evident that this was as near an approximation towards perfection as could possibly have been expected; and indeed much more than was likely to be effected in a large engine, as the vapour left beneath the piston possessed only onefifteenth part of the elastic force of the steam employed to form the vacuum.

Having discovered that the great waste of caloric in the old engine, arose from the alternate heating and cooling the cylinder, by the admission and subsequent condensation of the heated