ced by the foregoing observations; and it has been the want of a due attention to this fact that has puzzled and confused all the different authors on the cultivation and management of fruit trees. In making their comparisons, therefore, and in forming their maxims of practice, I must request my readers will bear those observations in mind.

New Experiments on some of the Combinations of Phosphorus. By Sir HUMPHRY DAVY, LL. D. F. R. S. V.P.R.I.

With an Engraving.

From the PHILOSOPHICAL TRANSACTIONS of the ROYAL SOCIETY of LONDON.

IN a paper, published in the Transactions of the Royal

Society, for 1812, I have detailed a number of experiments on phosphorus, from which I deduced the composition of some of its compounds with oxygen, with hydrogen, and with chlorine. Since the appearance of this paper, various researches have been brought forward on the same subject, in which some results, differing very much from each other, and from mine, are stated. I ventured to conclude, that the phosphoric acid contained double the quantity of oxygen to that in the phosphorous acid; and that phosphoric acid contained about three-fifths of its weight of oxygen.

M. Berzelius considers the oxygen in phosphoric acid to be 128.17, and M. Dulong, 124.5, the phosphorus being 100. M. Dulong and M. Berzelius suppose the quantity of oxygen in phosphorous acid to be to that in phosphoric acid as three to five.

The motive which immediately induced me to resume the enquiry respecting the phosphoric combinations, was

M. Dulong's paper. This ingenious chemist has not only endeavoured to establish new proportions in the known compounds of phosphorus, but has likewise attempted to prove the existence of two new acids of phosphorus; and has denied several facts which I considered as sufficiently established.

The details which I have to lay before the Society in the following pages, will serve to correct and fix, I hope, with tolerable accuracy, the proportional number or equivalent of phosphorus, and at the same time will show the truth of the general series of proportions that I assigned to its compounds. In a case where my conclusions differ so materially from those of M. Berzelius and Dulong, it may be supposed that I have not adopted them without considerable caution; and I have preferred my own results to theirs, only because they have been confirmed by minute and repeated experiments.

I was certain, from various experiments, made both long ago and recently, and the results of which had been confirmed by Mr. Brande, that the proportion of oxygen, which M. Dulong assigns to phosphoric acid, is considerably smaller than that denoted by the combustion of small quantities of phosphorus in oxygen gas. I knew that minute portions of phosphuretted hydrogen were separated from phosphorus by voltaic electricity; and it occurred to me as possible, that water might be formed in the combustion of phosphorus, and separated from the phosphoric acid when it entered into saline and metallic combinations. To ascertain if this were the case, I passed phosphorus to saturation through red hot lime in a green glass tube connected with a mercurio-pneumatic apparatus: the combination took place with vivid ignition; but no elastic fluid was produced. A portion of the phosphuret of lime formed, was introduced into a tray of

platinum,

platinum, and heated in a glass retort filled with oxygen gas; the phosphuret of lime burnt brilliantly, and became partly converted into phosphate of lime; but on restoring the original temperature of the retort, there was no appearance of vapour or of moisture.

Having examined the phosphate of lime formed in this operation, and satisfied myself that it was the same as that formed by other methods, it became evident that there were no sources of error in the experiments on the combustion of phosphorus in oxygen gas, arising from the formation or separation of water; and the only circumstance which could be urged against the accuracy of processes on this combustion, was the small quantity of materials * on which they had been made.

The vividness and rapidity of the combustion of phosphorus, renders it impossible to burn considerable quantities of phosphorus in the common way in glass vessels. Phosphuret of lime burns much more slowly and less in-tensely. I endeavoured to ascertain the quantity of oxygen absorbed by a given weight of phosphorus converted into phosphuret of lime; but the experiment did not succeed. Though the phosphuret of lime was in fine powder, and distributed over a large surface, yet the phosphate of lime which formed and fused on the exterior, defended the interior of the phosphuret from the action of the oxygen, and prevented its combustion.

After various unsuccessful trials to convert considerable quantities of phosphorus into phosphoric acid by combinations containing oxygen, I at last thought of a

* A source of error might be suspected in carbon combined with phosphorus; but I have been convinced, by experiments made on the action of chlorine on the phosphorus I employed, that it contained no appreciable quantity of carbon. I suspect that what is often taken for carburet of phosphorus, is in reality a red oxyd.

very simple mode of burning phosphorus, which answered perfectly.

Phosphorus requires a considerable heat for its volatilization. By inclosing it in a small tube, so constructed that the phosphorus can burn in vapour only from the aperture of the tube, large quantities of it may be burnt by the heat of a spirit lamp in a retort filled with oxygen, and the absorption of oxygen and the quantity of phosphoric acid formed may be minutely ascertained.

The accompanying sketch (Fig. 1, Plate IX.) will give an idea of the apparatus. The neck of the little curved tube, or small distilling retort, after the phosphorus is introduced, is drawn out, and an aperture left of about one-tenth of an inch; it should not be smaller, or it becomes choaked by the phosphoric acid formed. Regulating the heat by raising or lowering the spirit lamp, the combustion may be carried on slowly or rapidly at pleasure.

Operating in this way, I have often burnt from 5 to 10 grains of phosphorus without any accident, and ascertained exactly the quantity of oxygen absorbed : there is only one source of error-a quantity of phosphorus remains in the upper part of the tube, which cannot be burnt except by a greater heat than the retort will bear; and it is difficult to ascertain the precise weight of this, as the tube always unites with some phosphoric acid where it is red hot at its mouth; but this can be only a trifling source of error.

In these experiments, and in all the others detailed in this paper, I received much useful assistance from Mr. Faraday of the Royal Institution; and much of their value, if they shall be found to possess any, will be owing to his accuracy and steadiness of manipulation,

VOL. XXXIV.-SECOND SERIES. Pp Ex

<

Experiment I.

Six grains of phosphorus. The small tube with the phosphorus weighed before the combustion 56.5 grains; after the combustion 50.9; so that it had increased fourtenths; and this increase was in great measure from phosphorus that had escaped combustion; and when this was burnt out by a strong red heat, the increase of weight of the tube was under one-tenth; so that at least 5.9 of phosphorus had been converted into acid: 23.5 cubical inches of oxygen were absorbed; thermometer being at 46° Fahrenheit; barometer 29.6 inches.

Experiment II.

Ten grains of phosphorus. The glass tube containing the posphorus weighed 103.1 grains; after the experiment 95.6; but much phosphorus remained unconsumed. After the tube had been heated to redness, it weighed 94 grains; so that at least 8.4 grains of phosphorus were consumed in the first process. The absorption of gas was 34 cubical inches. Barometer 29.8, thermometer 47°.

Experiment III.

Ten grains of phosphorus. By weighing the tube after the experiment, and then distilling and burning the residual phosphorus, it was found that 9.1 grains of phosphorus had been burnt, which had absorbed 35.25 cubical inches of oxygen. Barometer 29.7, thermometer 49° Fahrenheit.

I give these experiments as the most accurate I have made. The pressure and temperature vary so little, that the corrections for them are of no importance. Supposing that 100 cubical inches of oxygen (the barometer being between 29.8 and 29.6, and the thermometer between 46° and 49° Fahrenheit) weigh 33.9 grains, phos