weight; but as it was not possible to weigh them except in the atmosphere, the moisture attracted rendered the result doubtful, and the proportions from the weight of the oxygen absorbed are more to be depended on. In the experiments in which the processes of wieghing were most speedily performed, and in which no alkali adhered to the tube, the basis of potash gained nearly two parts for ten, and that of the soda between three and four parts...

The results of the decompositions of water by the bases of the alkalies, were much more readily and perfectly obtained than those of their combustion.

To check the rapidity of the process, and in the case of potash to prevent any of the basis from being dissolved, I employed the amalgams with mercury. I used a known weight of the bases, and made the amalgams under naphtha, using about two parts of mercury in volume to one of basis.

In the first instances I placed the amalgams under tubes filled with naphtha, and inverted in glasses of naphtha, and slowly admitted water to the amalgam at the bottom of the glass; but this precaution I soon found unnecessary, for the action of the water was not so intense but that the hydrogen gas could be wholly collected.

I shall give an account of the most accurate experiments made on the decomposition of water by the bases of potash and soda. In an experiment on the basis of potash, conducted with every attention that I could pay to the minutia of the operation, hydrogen gas equal in volume to 298 grains of mercury, was disengaged by the action of .08 grains of the basis of potash, which had been amalgamated with about three grains of mercury. The thermometer at the end of the process indicated a temperature of 56° Fahrenheit, and the barometer an atmospheric pressure equal to 29.6 inches.

Now, this quantity of hydrogen* would require for its com

bustion a volume of oxygen gas about equal to that occupied by 154.9 grains of mercury, which gives the weight of oxygen to saturate the .08 grains of the basis of potash at the mean temperature and pressure, nearly .0151 grains; and .08+ .0151=.0951: .08 :: 100: 84.1 nearly.

And, according to these indications, 100 parts of potash consists of about 84 basis and 16 oxygen.

In an experiment on the decomposition of water by the basis of soda, the mercury in the barometer standing at 30.4 inches, and in the thermometer at 52° Fahrenheit, the volume of hydrogen gas evolved by the action of .054 grains of basis equalled that of 326 grains of quicksilver.

Now, this, at the mean temperature and pressure, would require for its conversion into water .0172 of oxygen, and .054+ .0172=.0712 : .054 :: 100: 76 nearly; and, according to these indications, 100 parts of soda consists of nearly 76 basis and 24 oxygen.

In another experiment made with very great care, .052 of the basis of soda were used, the mercury in the barometer was at 29.9 inches, and that in the thermometer at 58° Fahrenheit. The volume of hydrogen evolved was equal to that of 302 grains of mercury, which would demand for its saturation by combustion at the mean temperature and pressure, .01549 grains of oxygen; and 100 parts of soda, according to this proportion, would consist nearly of 77 basis and 23 oxygen.

The experiments which have been just detailed are those in which the largest quantities of materials were employed. I have compared their results, however, with the results of several others, in which the decomposition of water was performed with great care, but in which the proportion of the basis was still more minute. The largest quantity of oxygen indicated by these experiments was, for potash 17, and for soda 28 parts in 100, and the smallest 13 and 19; and, comparing all the estimations, it will, probably, be a good approximation to the truth to consider potash as composed of about six parts basis

and one of oxygen, and soda as consisting of seven basis and two oxygen.

VII. Some general Observations on the Relations of the Bases of Potash and Soda to other Bodies.

Should the bases of soda and potash be called metals? The greater number of philosophical persons to whom this question has been put have answered in the affirmative. They agree with metals in opacity, lustre, malleability, conducting powers as to heat and electricity, and in their qualities of chemical combination.

Their low specific gravity does not appear a sufficient reason for making them a new class; for amongst the metals themselves there are remarkable differences in this respect, platina being nearly four times as heavy as tellurium ;* and in philosophical division of the classes of bodies, the analogy between the greater number of properties must always be the foundation of arrangement.

On this idea, in naming the bases of potash and soda, it will be proper to adopt the termination, which, by common consent, has been applied to other newly discovered metals, and which, though originally Latin, is now naturalized in our language.

Potasium and sodium are the names by which I have ventured to call the two new substances; and whatever changes of theory, with regard to the composition of bodies, may hereafter take place, these terms can scarcely express an error; for they may be considered as implying simply the metals produced from potash and soda. I have consulted with many of the most

Tellurium is not much more than six times as heavy as the basis of soda. There is great reason to believe metals of a similar chemical nature to the basis of potash and soda will be found of intermediate specific gravities between them and the lightest of the common metals. Of this subject I shall treat again in the text in the course of the following pages.

eminent scientific persons in this country upon the methods of derivation, and the one I have adopted has been the one most generally approved. It is, perhaps, more significant than elegant; but it was not possible to found names upon specific properties not common to both; and though a name for the basis of soda might have been borrowed from the Greek, yet an analogous one could not have been applied to that of potash, for the ancients do not seem to have distinguished between the two alkalies.

The more caution is necessary in avoiding any theoretical expression in the terms, because the new electro-chemical phenomena that are daily becoming disclosed seem distinctly to show that the mature time for a complete generalization of chemical facts is yet far distant; and though, in the explanations of the various results of experiments that have been detailed, the antiphlogistic solution of the phenomena has been uniformly adopted, yet the motive for employing it has been rather a sense of its beauty and precision than a conviction of its permanency

and truth.

The discovery of the agency of the gases destroyed the hypothesis of Stahl. The knowledge of the powers and effects of the ethereal substances may, at a future time, possibly act a similar part with regard to the more refined and ingenious hypothesis of Lavoisier, but, in the present state of our knowledge, it appears the best approximation that has been made to a perfect logic of chemistry.

Whatever future changes may take place in theory, there seems, however, every reason to believe, that the metallic bases of the alkalies and the common metals will stand in the same arrangement of substances, and as yet we have no good reasons for assuming the compound nature of this class of bodies.*

* A phlogistic chemical theory might certainly be defended on the idea that the metals are compounds of certain unknown basis with the same matter as that existing in hydrogen; and the metallic oxyds, alkalies, and acids, compounds of the same basis with water; but in this theory more unknown princi

The experiments in which it is said that alkalies, metallic oxyds, and earths may be formed from air and water alone, in processes of vegetation, have been always made in an inconclusive manner;* for distilled water, as I have endeavoured to show,† may contain both saline and metallic impregnations, and the free atmosphere almost constantly holds in mechanical suspension solid substances of various kinds.

In the common processes of nature, all the products of living beings may be easily conceived to be elicited from known combinations of matter.

The compounds of iron, of the alkalies, and the earths, with mineral acids, generally abound in soils. From the decomposition of basaltic, porphyritic,‡ and granitic rocks, there is a con

ples would be assumed than in the generally received theory; it would be less elegant, and less distinct. In my first experiments on the distillation of the basis of potash, finding hydrogen generally produced, I was led to compare the phlogistic hypothesis with the new facts, and I found it fully adequate to the explanation. More delicate researches, however, afterwards proved, that in the cases when inflammable gases appeared, water, or some body in which hydrogen is admitted to exist, was present.

The explanation of Van Helmont of his fact of the production of earth in the growth of the willow was completely overturned by the researches of Woodward, Phil. Trans. vol. xxi. p. 193. The conclusions which Mr. Braconnot has very lately drawn from his ingenious experiments, Annales de Chimie Fevrier, 1807, p. 187, are rendered of little avail, in consequence of the circumstances stated in the text. In the only case of vegetation in which the free atmosphere was excluded, the seeds grew in white sand, which is stated to have been purified by washing in muriatic acid; but such a process was insufficient to deprive it of substances which might afford carbone, or various inflammable matters. Carbonaceous matter exists in several stones, which afford a whitish or greyish powder, and, when in a stone, the quantity of carbonate of lime is very small; in proportion to the other earthy ingredients it is scarcely acted on by acids.

† Bakerian Lecture, 1806, p. 8.

In the year 1804, for a particular purpose of geological inquiry, I made an analysis of the porcelain clay of St. Stephen's, in Cornwall, which results from the decomposition of feldtspar of fine-grained granite: I could not detect in it the smallest quantity of alkali. In making some experiments on speci