[1.97 in.] from its aperture. The second is a brass collar, in which the open end of the large glass tube is fixed by means of a cement, that will not fuse under 40° [104° F.] The third piece is a cock of a peculiar construction, in which all the merit of the apparatus consists. The key of this cock is solid, and may be turned into any position, without giving passage to the air; but about the middle of its length it has a superficial cavity, capable of holding a substance the size of a small pea. This cavity is so contrived, that when uppermost it answers to a small vertical funnel, which enters into the nozzle, and forms as it were its extremity; and when lowermost it communicates with the body of the cock, which is perforated, and screws into the brass collar before mentioned. Thus on putting small fragments of any thing into the funnel, and turning the key, the cavity is filled with them, and conveys them, on continuing to turn it, into the body of the cock, whence they fall into the brass collar, and so to the bottom of the glass tube.

tion.'

If this matter therefore be a mixture of some vegetable Its applica substance with hyperoximuriate of potash in suitable proportion, and if the lower part of the glass tube be sufficiently hot, it will scarcely touch it before it is vividly inflamed; when the vegetable substance will be instantaneously destroyed, and converted into water and carbonic acid, which may be collected over mercury, with the superfluous oxigen gas, by means of the small lateral tube.

To perform this operation readily, it is necessary, that Preliminary the matter should separate entirely from the cavity, and steps. fall to the bottom of the tube. For this purpose it is to be made into small balls, as will presently be described. It is Preparation of necessary too to inquire, what quantity of hyperoximuriate the substance to be analysed. will be sufficient for burning the vegetable substance completely; and at least half as much more must be used, that the combustion may be perfect.

But of all the preliminary steps the most important is the Analysis of analysis of the hyperoximuriate employed, for all the cal- the hyperoxiculations of the experiments are founded in great measure

on this analysis.

All this being well understood, it will be easy to conceive,

muriate.

scribed.

Process de. how a vegetable substance may be analysed with the hyperoximuriate. Let the substance to be analysed be carefully levigated, and let the hyperoximuriate be levigated separately weigh the quantity of each, dried at the heat of boiling water, in a very sensible balance; mix them intimately, moisten them, and mould them in cylinders; divide these cylinders into small portions, and round them between the fingers like pills; and lastly expose these to the temperature of boiling water for a sufficient time to render 'them as dry as the powders were before. If the substance to be analysed be a vegetable acid, it must be combined with lime or barytes, before it is mixed with the hyperoximuriate; the salt thus formed is to be analysed, and aecount taken of the carbonic acid that remains united with the base after the experiment; in fine, if the substance to be analysed contain any thing foreign to its nature, account must be taken of this also.

Proof of its accuracy.

Caution.

Analysis of the gasses.

Thus we know with precision, that a given weight of the mixture answers to a known weight of hyperoximuriate and the substance to be analysed.

Now, to finish the operation, all that is required is, to bring the bottom of the tube to a cherry red heat; to expel all the air by means of a certain number of balls, which need not be weighed, and which are dropped into it one after another; and then to decompose a quantity accurately weighed, and carefully collect all the gasses in phials filled with mercury, and previously measured.

If all the phials be of the same size, they will be filled with gas by equal weights of the mixture; and if the gas be examined, it will be found precisely similar, an evident proof of the extreme accuracy of this mode of analysis.

During the whole of the process the tube should be kept at the highest degree of heat it can support without fusion, that the gasses may contain no oxicarburetted hidrogen, or as little as possible. In all cases the analysis should be made over mercury. This is a trial which is indispensable. It is sufficient to mix them with a fourth of their bulk of hidrogen, and to take the electric spark in them. As they include a great excess of oxigen, the hidrogen added, of which account must be taken, burns as well as all the oxicarbu

retted

retted hidrogen they may contain; and thus we acquire a certainty, that they no longer consist of any thing but carbonic acid and oxigen, the separation of which is to be effected by means of potash.

cautions.

But this necessity of raising the temperature so high, Farther preobliges us, on the other hand, to take some precautions for preventing the cork from being heated. For this purpose the glass tube is passed through a brick, into which it is luted with clay, which has the advantage, at the same time, of rendering the apparatus firm; and besides, a small hollow cylinder is soldered to the body of the cock, to contain water, or ice, which is still better.

Thus we have all the necessary data for knowing the proportion of the principles of the vegetable substance. We know how much of it has been burned, for we have its weight to half a milligramme, [about eight thousandths of a grain]; we know how much oxigen was required to convert it into water and carbonic acid, since the quantity is the dif ference between that contained in the hyperoximuriate and that found in the gasses produced: lastly, we know how much carbonic acid has been formed, and can calculate how much water must have been produced,

Data.

stances,

By following the same method of analysis, we may equally Analysis of determine the proportions of the constituent principles of anal sub all animal substances. But as these substances contain nitrogen; and nitrous acid gas would be formed, if an excess of hyperoximuriate were employed for burning them; only such a quantity must be used, as is sufficient to reduce them completely to carbonic acid gas, oxicarburetted hidrogen, and nitrogen, which are to be analysed in the mercurial eudiometer by the common methods, whence we deduce with precision the proportions of the principles of the animal substance itself.

The mode in which we proceed in the analysis of vegetable Small quanti and animal substances being exactly known, we may say the results ac ties used, but what is the quantity we decompose, without fear of dimi- curate. nishing the reliance, that may be placed on our results. This quantity extended, at most, to 6 dec. [9.27 grs.] If, however, the least doubt should arise respecting their extreme accuracy, we should remove it by observing, that we

tus and of the

method.

filled with gas two and sometimes three phials of the same size in succession, that these gasses were absolutely the same, and always came from the same weight of the sub

stance.

Accuracy of an We may add, that the precision of an analysis depends analysis de much more on the accuracy of the instruments, and of the pends chiefly on the nicety methods employed, than in the quantity of the substance on of the appara which we operate. The analysis of air is more accurate than any analysis of salts, though it is made on two or three hundred times less matter; because in the former, where we judge of weights by very considerable bulks, the errours to which we are liable are perhaps ten or twelve hundred times less sensible than in the second, where we have not this resource. Now, as we convert into gas the substances we analyse, we bring our analyses not merely to the certainty of ordinary mineral analyses, but to that of mineral analyses of the greatest accuracy; particularly as we collect at least a quart of gas, aud in our method of proceeding itself find the proof of an extreme accuracy, and of the most trifling

errours.

Vegetablesub By this method, and with all the precautions we have stances already mentioned, we have already analysed sixteen vegetable subanalysed. stances; namely, the oxalic, tartarous, mucous, citric, and acetic acids; yellow resin, copal, wax, and olive oil; sugar, gum, starch, sugar of milk, beech wood, oak, and the crystallizable principle of manna. The results we have obtained seem to us highly interesting, for they have led us to three remarkable laws, to which the composition of vege tables is subjected, and which may be expressed as follows. 1. A vegetable substance is always acid, whenever its oxigen is in greater proportion to its hidrogen than would form

Laws of vegetable composi tion.

water,

2. A vegetable substance is always resinous, or oily, or alcoholic, &c., whenever its oxigen is in smaller proportion to its hidrogen than would form water,

3. Lastly, a vegetable substance is neither acid, nor resinous, but analogous to sugar, gem, starch, sugar of milk, woody fibre, or the crystallizable principle of manna, whenever its oxigen is in the same proportion to its hidrogen as. would form water,

Thus,

Thus, if we were to suppose, for a moment, that the hi- Vegetable drogen and oxigen were in the state of water in vegetable acids; substances, which we are far from considering as true, vegetable acids would be formed of carbon, water, and oxigen,

in different proportions:

Resins, fixed and volatile oils, alcohol, and ether, would resins, &c. be formed of carbon, water, and hidrogen, also in different proportions and

Lastly, sugar, gum, starch, sugar of milk, woody fibre, sugar, &c. and the crystallizable principle of manna, would be formed of carbon and water alone, and would differ only by the greater or less quantity they contained.

This we may show by quoting various analyses of acid and resinous substances, and of substances that are neither acid nor resinous.

Oxalic acid therefore contains more than half its weight These the two of oxigen in excess with respect to its hidrogen; while in extremes, acetic acid this excess is not quite three hundredths.

These two acids occupy the extremities of the series of

vegetable acids: one is the most oxigenized of them, the

facts.

other the least. This is the reason why so much nitric acid Explanation is required to convert sugar, gum, &c., into oxalic acid; of certain why, on the contrary, many vegetable and animal substances so easily produce acetic acid in a number of instances; and why, in particular, wine is changed into vinegar without

the