MESSRS. EDITORS—At the close of an article upon heat and cold in your issue of October 31st, you express a desire to hear more upon the subject of heating water hy friction, which has called out an article from Joseph E. Holmes, Esq., and the result at which he arrives for observing the action of the Gwynne pump (and in which you appear to coincide) is that owing to a large suction pipe being used, and the water forced through a smaller opening, a compression of the particles of air contained in the water ensues, that results in evolving heat, which, being taken up by the water, raises its temperature. My experience leads me to a different opinion. I have had running, during the late Fair of the American Institute, several centrifugal pumps; one had a suction pipe of 5 inches diameter, with a discharge 2 inches, through which, at an elevation of some 3 feet, 750 gallons of water per minute were forced; and the pump was kept running during the whole time the Fair was kept open, whenever the engines were operating, pumping the same water without change from first to last, except to replace the loss from evaporation, and At no time did the tenppfcratur approach blood heat, though perceptibly tightened after several hours' constant running. I had also another pump, with a 9-inch suction, and 6-inch discharge pipe, elevating its water about 5 feet, and discharging 3,000 gallons per minute, running for three weeks, under the same circumstances as the first, which did not raise the temperature to the same extent, although the whole quantity of water used was less, in proportion to the quantity discharged, than in the first instance. These facts, in my opinion, would show that the increased heat was due to the increased friction of small pipes. I account for the different results in Mr. Holmes' experience and my own as follows:— The discharge pipe, in the case he instances, being smaller, and discharging about the same proportionate quantity, caused an increased friction, and tne angles at which the water passes through the Gwynne pump causes a large increase of friction—there being no angles in my passages, the water being passed around curves, which become easier as the size of the pump is enlarged. To these causes' may, perhaps, be added, as you suggest, an increase of temperature from the friction of the parts of the pump, as I find, by the manufacturers' published tables, that to raise water 15 feet with the Gwynne pump, through a 4 inch suction, and 2J-inch discharge pipe, at the rate of 100 gallons per minute, requires a velocity at the periphery of the rotating wheel of 3,750 feet per minute; while with a 3-inch suction and 2-inch discharge pipe, my pump discharges 75 gallons per minute, 17 feet high, with the periphery of the wheel running less than 2,000 feet per minute. I should be pleased to hear the opinion of yourselves, or oome of your scientific correspondents, upon the facts as stated. W. D. ANDREWS. New York, November 27, 1857. [The fact seems to be established, that the water which passes through centrifugal pumps is elevated in temperature. What is the cause of this ? is the question. The water in these pumps, it seems, is not compressed, but is driven at a high velocity. If the air in the water is not compressed, then the rise in temperature cannot be due to this cause. Is it due to the friction of water on the metal? This idea is opposed to that hitherto entertained respecting the friction of fluids on solids. There is no question respecting the fact that the friction of the solid parts of these pumps generates frictional heat, which must be car- L' by the water ; but Mr. Andrews is of opinion that the friction of the water on the metal also, generates heat. This is an interesting point, or rather question; for if it be established that tne Mfctio of fluids on solids gen- erates heat, then the fcemperateeof the water discharged by such pumps will afford a very good test of their efficiency, as the one which raises the temperature of the water highest must require the greatest amount of power to operate, and thus be the least effective.—EDS.