276] F I L the foot of a mountain. The cause of this difference appears to be ow- ing to the circumstance, that if the ■water only descends through sand, the finest and most weighty fo- xeign particles gradually penetrate through the sandy strata ; on the contrary, when it is forced to rise through sand, all such ponderous ingredients settle at the bottom ; because, from their greater specific gravity, they cannot ascend to the top. The lighter particles of fluids, consequently, in both cases remain in the upper strata of the earth or sand. From these considerations, Pro- fessor Parrot, jun. of Paris, was induced to give his filtering ma- chine the form represented in the following cut : The principal part of the ma- chinery consists of a square vessel, bent in the form of an inverted syphon. The curve may be cir- cular, elliptic, or in any other di- rection.' This vessel is filled with fine, pure sand, till nearly the height of the dotted line x, y, which denotes the ascent of the water to D, whence it flows into the receiver. The part marked A, B, should always project above FIL this line, according to the size of the filtering machine. To A, B, there is attached a woollen bag, which is open at the top, and the lower part of which touches the sand. It serves the purpose of col- lecting the coarsest impurities, and thus preserves the sand for a longer time from becoming foul. The bag, therefore, may occasionally be removed, and rinsed in clean water. — It is evident, that the wa- ter flows at A, through the bag into the filtre, and rises at the place marled D, which is considerably lower than the former. It affords a very agreeable sight to observe the most limpid fluid penetrating the uppermost stratum of sand, pei fecfiy similar to that oozing from the purest natural spring. Prof. Parrot remarks that he procured a filtering machine made of block-tin, for ascertaining by experiments the purity and quality of water, that may thus be obtain- ed in a given time. It consisted of the following dimensions : the small diameter B,E, was eight Paris inches ; the large of the whole machine, eleven inches ; conse- quently the thickness of the. ves- sel A, B, was one inch and a half; — the breadth of it, two inches and seven-eighths. The perpen- dicular height of the lower side, from C, its basis, to the rim D, whence the water issues, was four inches and one-twelfth — the oppo- site height of the mouth A,B, eight inches and three-fourths; and thu height of the sand on the side marked D, was three inches and one- sixth. Although, in experiments of this nature, much depends on the re- lative size and purity of the sand, which necessarily afford different results, yet Prof. Parrot has, af- ter repeated trials, deduced the follow-
