During the course of the enquiry, Davy is led to suppose that air may exist in mercury, in the same invisible state as it does in water, that is, distributed through its pores; and that absorption of air may, therefore, explain the difference of the heights of the mercury in different barometers. This, it must be confessed, if true, is a most disheartening fact, as it at once precludes the possibility of any thing like accuracy in our barometers; but Mr. Daniell, to whom on all subjects of meteorology we are bound to pay the greatest deference, differs altogether from our philosopher upon this point, and he adduces a single observation which he thinks nearly disproves the supposition. "All fluids," says he, "which are known to absorb air into their pores, invariably emit it when the pressure of the atmosphere is removed; but, upon an extensive examination of large bodies of mercury, variously heated in the vacuum of an air-pump, I never saw a bubble of air given off from the surface of the metal." Davy, it must be stated, obtained a far different result; but an observation of Mr. Daniell explains the cause of it. "Air," he continues, "will rise from the contact of the mercury with the glass in which it is contained, in exact inverse proportion to the care with which it has been filled, but it _never rises from the surface of the mercury alone_. The difficulty of properly filling a barometer tube, I attribute to the attraction between the glass and the air--not to that between the mercury and the air."[77] [77] "Meteorological Essays and Observations," p. 363.--See also Bellani's experiments upon this subject, which are so satisfactory as to remove every doubt from the subject. On the 13th of June 1822, a memoir was read before the Royal Society, "On the state of Water and Aëriform matter in cavities found in certain Crystals. By Sir Humphry Davy, Bart. P.R.S." It is generally admitted by Geologists, that the greater number of the crystalline substances of the mineral kingdom must have been previously in a liquid state; but different schools have assumed different causes for their solution; some attributing the effect principally to the agency of water, others to that of heat. In the paper under consideration, the author very freely avows himself as the champion of the latter doctrine. "When it is considered," says he, "that the solvent power of water depends upon its temperature, and its deposition of solid matters upon its change of state or of temperature, and that, being a gravitating substance, the same quantity must always belong to the globe, it becomes difficult to allow much weight to the arguments of the Wernerians, or Neptunists, who have generally neglected, in their speculations, the laws of chemical attraction.