foon as poffible, well corked, and cemented. It will keep, however, very well, if the bottle be only well corked, and kept with the mouth downwards.

Obfervations.

1. The bafon may be placed inverted upon the veffel full of water, with a flip of paper between them, and then both turned upfide down together; but all this trouble will be faved by having a larger veffel of water, in which both of them may be immersed.

2. If the veffel containing the water to be agitated be large, it may be most convenient first to place it inverted, in a bafon full of water, and then to draw out the common air by means of a fyphon, either making ufe of a fyringe, or drawing it out with the mouth. In this cafe, also, some kind of handle fhould be fastened to the bottom of the veffel, for the more eafy agitation of it.

3. A narrow mouthed veffel is not neceffary, but it is the moft proper for the purpose, because it may be agitated with lefs danger of the common air getting into it.

4. The flexible pipe is not neceffary, though I think it is exceedingly convenient. When It is not used, a bent tube, a, fig. 2. (for which glafs is the most proper) must be ready to be inferted into the hole made in the cork, when the bladder containing the fixed air is separated from the phial, in which it was generated, The extremity of this tube being put under the veffel of water, and the bladder being com preffed, the air will be conveyed into it, as before.

i

5. If the use of a bladder be objected to, though nothing can be more inoffenfive, the phial containing the chalk and water must not be agitated at all, or with the greatest caution unless a finall phial, a, fig. 3. be interpofed between the phial and the veffel of water, in the manner reprefented in the drawing. For by this means the chalk and water that may be thrown up the tube b will lodge at the bottom of the phial a, while nothing but the air will get into the pipe c, and fo enter the water. If the tube b be made of tin or copper, the fmall phial a will not need any other fupport, the cork into which the extremities of both the tubes are inferted being made to fit the phial very exactly.

6. The

6. The phial e, fig. 1. fhould always be placed, or held, confiderably lower than the veffel a; that if any part of the mixture fhould be thrown up into the bladder, it may remain in the lower part of it, from which it may be eafily preffed back again. This, however, is not neceffary, fince if it remain in, the lower part of the bladder, nothing but the pure air will get into the pipe, and fo into the water,

7. If much more than half of the veffel be filled with air, there will not be a body of water fufficient to agitate, and the process will take up much more time.

8. If the chalk be too finely powdered, it will yield the fixed air too faft.

9. After every procefs, the water to which the chalk is put must be changed.

10. It will be proper to fill the bladder with water once every day, after it has been used, that any of the oil of vitriol which may haye got into it, and would be in danger of corroding it, may be thoroughly diluted.:

11. The veffel, which I have generally made ufe of, holds about three pints, and the phial containing the chalk and water is one of ten

ounces;

ounces; and I find that a little more than a tea-fpoonful of oil of vitriol is fufficient to produce as much air as will impregnate that quantity of water.

12. If the veffel containing the water be larger, the phial containing the chalk and the oil of vitriol fhould either be larger in proportion, or fresh water and oil of vitriol muft be put to the chalk, to produce the requisite quantity of air.

13. In general, the whole process does not take up more than about a quarter of an hour, the agitation not five minutes; and in nearly the fame time might a veffel of water, containing two or three gallons, or indeed any quantity that a perfon could well shake, be impregnated with fixed air, if the phial containing the chalk and oil of vitriol, be larger in the fame proportion.

14. To give the water as much air as it can receive in this way, the process may be repeated with the water thus impregnated. I generally chufe to do it two or three times, but very little will be gained by repeating it oftener; fince, after fome time, as much fixed air will escape from that part of the surface of the

water which is exposed to the common air, as can be imbibed from within the veffel.

15. All calcareous fubftances contain fixed. air, and any acids may be used in order to fet it loose from them; but chalk and oil of vitriol are, both of them, the cheapest, and, upon the whole, the best for the purpose.

16. It may poffibly be imagined that part of the oil of vitriol is rendered volatile in this process, and so becomes mixed with the water; but it does not appear, by the most rigid chymical examination, that the leaft perceivable quantity of the acid gets into the water in this way; and if fo fmall a quantity as a fingle drop of oil of vitriol be mixed with a pint of water (and a much greater quantity would be far from making it lefs wholefome) it might be discovered. The experiments which were made to afcertain this fact were made with diftilled water, the difagreeable taste of which is not taken off, in any degree, by the mixture of fixed air. Otherwife, diftilled water, being clogged with no foreign principle, wil imbibe fixed air fafter, and retain a greater quantity of it than other water. In the experiments that were made for this purpofe, I was affifted by Mr. Hey, a furgeon in Leeds, who is well

fkilled