But the triumph of Paical was too complete not to excite afresh the malignity of his enemies : Descartes, among others, attempted to deprive him of the honour of the discovery of the pressure of the air; and iu a letter which he wrote to M. Carcavi, dated June 11. 1649, accuses Pascal of a waist of candour, and asserts that lie first suggested to him the experiment of Puy de-dome. Pascal, on the other hand, maintains, in the most solemn manner, that the experiment was en tirely his own, and that he never received the smallest hint of it from any person. It would be improper to enter here into the merits of the dispute ; but it appears to us, that the pretensions of Descartes are al-. together groundless, and that the discovery of atmo-. spherical pressure is clue to Pascal alone. ' It is obvious, from the experiment of Puy-de dome, that the Torricellian tube, if properly gra duated, may be employed to measure the heights of mountains; a purpose, we already mentioned, to which it is now frequently applied. Boyle suggested this application of it in 1665.
After it was ascertained, that the weight of the air was the true cause of the suspension of the mer cury in the tube of Torricelli, Perrier continued to make daily observations with it, from the beginning of 1649 to the end of March 1651. One of his friends at Paris, and Descartes, who was then at Stockholm, made similar observations during the same period, and they found, that the column of mercury varied in length, according to the tempera ture, the winds, the moisture, and other circumstan ces with the state of the atmosphere. Thus the tube of Torricelli became an instrument, not only for shewing the weight of the air, but for pointing out the changes of weather which lizppeu in consequence of variations in the weights of the at mospherical columns. Some pretend that Otto de Guericke first proposed to employ the barometer as a weather-glass ; but this idea, so obvious, was a natural consequence of the observations of Perrier •and Descartes, and must have occurred at a very ear ly period to these philosophers. Boyle made many meteorological observations with it in 1666 ; and he was also at great pains to refute a hypothesis advan ced by Linus, concerning the cause of the elevation of the mercury in the Torricellian tube. This hy pothesis, which prevailed for some time,. was called the Funicular hypothesis: it assumed, that the mer cury was supported by an invisible rope of the same me tallic fluid; an opinion so extremely absurd, that, in the present state of science, we have some difficulty in believing that it was ever maintained, or that it should have been deemed worthy of a grave refuta The experiment of Torrieelli was so simple, and yet •so easy to be exhibited under a variety of forms, that a great number of barometers were soon proposed, either with a view of rendering them more correct, or enlarging the extent of the barometrical scale.
Before we proceed to give particular descriptions of these instruments, and of the various attempts which have been made to increase their accuracy and sensi bility, it may not be improper to make some previous •remarks applicable to barometers in The tubes intended for barometers ought to be sealed hermetically at both ends, immediately after . they are made at the glass-house, and to be kept in that state till they are to be fitted up. Without this precaution, they are apt to be sullied with dust, moisture, and other impurities, which it is almost impossible afterwards to remove, on account of the smallness of their diameters. When they are opened, which may be done with a file, care should be taken not to breathe into them, or to wash them with spi rit of wine, or any other fluid ; experience having proved, that in tubes so treated, the mercury al ways stands a little below its proper level. This is, no doubt, owing to the adhesion of a little of the spirit of wine to the sides of the tube, which being afterwards converted into vapour, renders the vacuum above the mercury imperfect. If any cleaning is ne cessary, it may be done with a fine linen rag, that has previously been well dried.
The tubes ought to be as perfectly cylindrical as possible, though, in some cases, this is not absolute ly necessary. They should be about 33 inches in length, and the diameter of their bore should be at least 2 or 21 lines, otherwise the friction, and capil lary action, will be apt to affect the free motion of the mercury. The 'glass should not be very thick, as it is apt, in that case, to break, when the mercury is boiled in the tube : half a line is sufficient.
The mercury ought to be perfectly pure, and free from all foreign metals. The best is what has been recently revived from cinnabar ; the common mer cury of the shops being often adulterated intention ally with tin, lead, bismuth, stands at various heights in the tuhe, according to the nature and quantity of the foreign substances with which it is amalgamated.
The different mechanical methods which have been proposed for purifying mercury, arc, for the most part, ineffectual ; we would, therefore, recommend the revivification of the metal from cinnabar, for nice barometers, as being least liable to uncertainty. For this purpose, take a pound of cinnabar, and reduce it to powder ; mix it well with five or six ounces of iron, or steel filings ; and having put the mixture in to an iron retort, expose the whole to the heat of a reverberatory furnace ; the mercury will soon pass over in a state of great purity, and may be obtained, by adapting to the retort an earthen receiver which has been previously half filled with water.