duced through the funnel G, the lower orifice of which is opened and closed by a screw worked by the small wheel 0. A piston, worked by the large wheel, compresses the water in the cylinder. The water in the cylin

FIG. 3.-CAILLETET'S LARGE APPARATUS FOR LIQUEFYING GASES.

A, Screw-press for compression; m, flint-glass cylinder inclosing the glass tube in which the gas is liquefied.

der passes through capillary metallic tubes to manometers (to show the pressure) and to a reservoir of mercury a, which is forced up into the glass tube m. This glass tube, which holds the gas to be liquefied, is surrounded by another containing a freezing mixture, and the whole is covered by a glass shade, on the floor of which is placed some substance which has strong affinity for moisture, and which prevents the deposit of vapor on the outside of the tube, hindering observation. The high pressure of the water forces the mercury up into the tube, thus compressing the gas. "If now," says M. Cailletet, "we inclose oxygen or pure carbonic oxide in the compression-apparatus; if we reduce these gases to a temperature of -29° Cent. by the aid of sulphurous acid at a pressure of about 300 atmospheres, both gases still retain their gaseous state. But if they be released suddenly [by reversing the small wheel O], so, according to Poisson's formula, producing a temperature of at least 200° below the starting-point, we at once see a heavy mist, caused by the liquefaction, or even, perhaps, the solidification, of the oxygen or carbonic oxide. The same phenomenon is observed in releasing carbonic acid and protoxide and bioxide of nitrogen, which have been subjected to strong pressure."

After having obtained these results, at a session of the Academy on December 31st, M. Cailletet announced that he had won a complete victory over the other permanent gases. M. Dumas informed the members present at the session that the able experimenter had succeeded in liquefying nitrogen, atmospheric air, even hydrogen itself, which would seem to have been the most refractory gas of them all. The New Metals Davyum and Neptunium.M. Sergius Kern, of St. Petersburg, discovered, toward the middle of the year 1877, a new metal belonging to the platinum group, to which he gave the name of Davyum, in honor of Sir Humphry Davy. Dissolved in aqua regia, and treated with potassa, davyuın yields a yellow precipitate, hydrate of davyum. Chloride of davyum, dissolved in a solution of potassic cyanide, yields, in crystals, a double cyanide of davyum and potassium. A concentrated solution of davyum chloride, with potassic sulphocyanide, gives a red precipitate, which, on being slowly cooled, yields large red crystals; if this precipitate be calcined, the sulphocyanureted davyum assumes the form of black powder. Davyum chloride forms double salts with chlorides of potassium and ammonium; these are insoluble in water, but highly soluble in absolute alcohol. The double salt of sodium and davyum is nearly insoluble in water and alcohol. Three experiments made to determine the density of davyum yielded, at temperature 24° Cent., these results, namely: 9.383, 9.387, 9.392. The author is of the opinion that the atomic weight of davyum is over 100-probably about 150 to 154.

Another new metal, discovered during the

An

past year, is Neptunium, found by Hermann in a mineral coming from Haddam, Conn. The history of this discovery is briefly stated as follows in the American Journal of Science, which publishes a synopsis of a communication from the discoverer to a German scientific journal. The mineral worked on was labeled "tantalite," but, on examination, it proved to be columbite and ferroilmenite in equal parts. The metallic oxides separated from the mineral consisted of Ta2Os 32.39, Cb,O, 36.79, IO, 24.52, Np.O, 6.30. To obtain the neptunium, the pulverized mineral was fused with hydropotassium sulphate, the acid hydrates digested with ammonium sulphide and hydrochloric acid, washed well with water, dissolved in hydrofluoric acid, mixed with an equivalent quantity of potassium fluoride, and the solution diluted to 40 parts boiling water to one of fluoride. On cooling, tantalum-potassium fluoride crystallized in delicate prisms. On evaporation, columbium-potassium fluoride and ilmenium potassium fluoride crystallized out, leaving an acid mother-liquid. This was di luted with 20 parts water, heated to boiling, and sodium hydrate added in excess. amorphous precipitate of sodium neptunate was formed, mixed with minute crystals of columbate. The precipitate was collected on a filter, pressed out, and boiled with 25 parts of water. The columbate dissolved, the neptunate remained. The latter was fused with hydro-potassium sulphate, the fusion was treated with boiling water, and the undissolved residue of neptunic acid washed and dried over sulphuric acid. Neptunic acid resembles in general the other acids of the group, but is distinguished from columbic and ilmenic acids by the insolubility of the sodium double fluoride, and from tantalic acid by the ready solubility of its potassium double fluoride. Neptunic acid gives with phosphorus salt in the inner blow-pipe flame a wine-yellow bead, the sodium salt a gold-yellow glass; while tantalic acid gives no color, columbic acid gives blue, and ilmenic acid gives brown. With tincture of galls, the sodium salts give, on addition of hydrochloric acid, a sulphur-yellow precipitate with tantalic, orange with columbic, brick-red with ilmenic, and cinnamon-brown with neptunic acid. Boiled with tin and hydrochloric acid, neptunic acid gives, like columbic and ilmenic acids, a blue solution. From the pure crystallized double potassium fluoride, the atomic weight of neptunium was fixed.as 118, its atomic volume as 18, and its specific gravity as 6.55. The formula of the acid is Np、O7, (H2O)15. The sodium salt crystallizes in prisms. The author prepared metallic columbium and ilmenium in the pure form, and determined the amount of oxygen taken up by these metals on heating them in the air. Columbium required 20.49 and ilmenium 37.96 of oxygen; the amount obtained by Rose being 20.60, and by Marignac 38.00. Rose, therefore, it is clear, had pure columbium; while it is equally clear, according

to Hermann, that Marignac must have had nearly pure ilmenium. This is stated to be a necessary result from the method of preparation. After crystallizing out the tantalumpotassium fluoride, Marignac evaporated and recrystallized, obtaining a nearly pure ilmenium-potassium fluoride, from which he prepared his metal. Hermann's paper concludes with an account of his methods of separating the metals of this group, and descriptions of their compounds.

Sensitiveness of Silver Salts.-In continuation of his researches on the sensitiveness of silver salts (American Journal of Science and Arts, No. lxxvii.), M. Carey Lea recognizes three modes in which salts of silver may exhibit their sensitiveness to light, viz.: they may exhibit a visible darkening; or they may receive a latent image, and this may have a capacity of being rendered visible either by receiving a deposit of metallic silver, or by decomposition by alkalies in connection with reducing agents. In the former of these two last-mentioned modes, the image is produced entirely by the addition of silver not previously present; in the latter, no silver whatever is added, but that portion of substance which received the direct action

of light undergoes decomposition by subsequent treatment. In both cases molecular change is set up by the action of light: the portions acted upon by light become, in the one case, more apt to attract a precipitate in the act of formation; in the other case they are more readily attacked by certain reducing agents. Now, while the silver compounds which exhibit the greatest tendency to form latent images by the action of light are the iodide, bromide, and chloride, Mr. Carey Lea finds that the same tendency is shared, though to a less degree, by other compounds, and that the latent images formed upon them may belong to either of the above-mentioned classes. In making his experiments, the author selected soluble salts of acids capable of forming insoluble or nearly insoluble salts with silver, and with them he impregnated the surface of very pure paper. After drying, the papers were floated on a solution of silver nitrate containing about 20 grains to the ounce, acidulated with half a drop of nitric acid (specific gravity 1.28), to the ounce of solution. The excess of silver nitrate having been worked out, one set of papers were then simply dried, and another set were soaked about a minute in a 10-grain solution of gallo-tanic acid, and then washed again. The salts thus formed on the paper were exposed to a strong diffuse light, some for 7, some for 12 seconds. They were next submitted to the action of a very weak solution of pyrogallol, ammonium carbonate, and potassium bromide, the latter being used to check the rapidity of the action of the other agents.

The results were as follows:

Silver citrate and tartrate both gave rather weak images. The citrate showed a strong tendency to irregular reduction. Nothing of this appeared in the case of the tartrate.

Silver platinocyanide gave quite a strong imagestronger than any other substance tried, except, of course, the silver bromide used for comparison. Silver mucate gave a very faint image with much irregular reduction.

Silver pyrophosphate behaved in the same way. Silver arsenite gave a moderately strong image, and free from all irregular reduction. coming next to the platinocyanide, and, like it, clear

Silver sulphocyanide, an extremely faint image with much irregular reduction.

Silver antimonio-tartrate, a weak image entirely free from irregular action.

but also clear. Silver fulminurate, weaker than the last mentioned,

Silver nitrate, similar to the last.

Silver hippurate, an excessively faint image with much irregular reduction.

The following substances showed (with the above-mentioned exposures) no trace of a latent image:

As respects the action of tannin, which was above salts, it appeared that no substance inseparately investigated with every one of the sensitive in the absence of tannin acquired sensitiveness by its presence. It was also doubtful if in any case tannin increased the sensitiveness of any of these substances-a fact which, in view of the increased sensitiveness conferred by tannin on the silver baloids, is remarkable.

New Acids. A new acid of phosphorus and Oxygen, standing between phosphorous and Salzer, of Worms. According to the old notaphosphoric acid, has been discovered by Th. tion, this acid, which has been named hypophosphoric acid, consists of 1 atom of phosphorus and 4 atoms of oxygen. rather insoluble salt. Salzer finds that the acide phosphatique of Pelletier is a mixture of phosphorous and hypophosphoric acids.

It forms a

C. Stahlschmidt has discovered a new organic acid, polyporic, occurring in certain the stems of diseased or dead oaks. The emfungi of the family polyporus, which grow on pirical formula is C.H7O2. This acid has a yellow color, and is so completely insoluble in water that the slightest trace of a soluble polythe addition of salt or of sulphuric acid. In porate in water renders the liquid turbid on virtue of this property the soluble polypothe turbidity serving instead of the usual change rates may act as indicators in alkalimetry, of color. With all bases it forms well-defined salts, of which the soluble ones, those of the alkalies, form deep-purple solutions. On heatcombustion-tube along with zinc-powder, bening polyporate of potassium to redness in a its conversion into nitro-benzol. zol was obtained, which was identified by

In a paper on the "Chemistry of Cocoa Butter," Mr. C. T. Kingzett described two new

fatty acids, prepared by saponifying the butter, and decomposing the soaps with dilute sulphuric or hydrochloric acid; they were purified by recrystallization from alcohol, fractionating, etc. The new acids are represented by the formula C12H24O2 and C84H128O2. The first is the formula of lauric acid, but it melts at 57.5° (lauric acid melting at 43° Cent.), so it must contain some acid of a higher melting-point than lauric acid, and therefore the acid itself must be lower in the series CH2O2 than lauric acid. The highest known acid in this series is melissic acid, Cao HooO2; the new acid has a formula not lower than C6H128O2. The lower acid crystallizes in pearly plates or fine long needles. The higher acid-for which the author proposes the name of "Theobromic Acid" -crystallizes in microscopic needles or granules, melts at 72.2° Cent., at a high temperature distills apparently unchanged, and is somewhat electric when dry-a property which is possessed in a high degree by its silver salt. The total fatty acids of cocoa-butter contain about 20 per cent. oleic acid. The author, in conclusion, points out that text-books state that "cocoa butter yields, almost exclusively, stearic acid." From the present investigations it is clear that this statement is incorrect. It is based entirely on determinations of the melting-point of the fatty acids obtained.

Estimation of Alcohol in a Watery Mixture. -Dr. Werner Siemens has contrived an ingenious apparatus, by which a stream composed of alcohol and water, mixed in any proportion, is so measured that a train of counter-wheels records the volume of the mixture, while a secured counter gives a true record of the amount of alcohol contained in it. The modus operandi is described as follows:

The volume of liquid is passed through a revolving drum, divided into three compartments by radial divisions, and not dissimilar in appearance to an ordinary wet gas-meter. The revolutions of this drum produce a record of the total volume of passing liquid. The liquid on its way to the measuring-drum passes through a receiver containing a float of thin metal filled with proof-spirit, which float is partially supported by means of a carefully-adjusted spring, and its position determines that of a lever, the angular position of which causes the alcohol-counter to rotate more or less for every revolution of the measuring-drum. Thus, if water only passes through the apparatus, the lever stands at its lowest position, and then the rotative motion is not communicated to the alcohol-counter, and this motion is rendered strictly proportionate to the alcohol contained in the liquid, allowance being made in the instrument for the change of volume due to chemical affinity between the two liquids.

Determination of Copper.-A new method of determining very small quantities of copper is offered by J. M. Merrick, of Boston. It is intended as a supplement to Bergeron and L'Hôte's colorimetric test, which fails to indicate a quantity of copper less than 0.5 milligramme. Mr. Merrick's method consists simply in concentrating to a very small bulk the solution suspected to contain copper, and

then depositing the copper, if present, upon platinum, by the battery. He uses for a depositing-cell a very small test-tube, on a foot cut off so as to give a vessel about 1 inch deep. Into this is introduced the solution acidified with sulphuric acid, and a platinum anode and cathode-each about an inch long and one-eighth of an inch or less wideare hung face to face, and very close together; and, the circuit being completed, very satisfactory deposits of copper are obtained, with incredibly minute quantities of the metal. The amounts are determined by the increased weight of the cathode (which is provided with a platinum wire soldered on with gold, by which it can be hooked to a balance), and on the loss of weight of the same after washing with nitric acid. The platinum is polished and heated red-hot before the first weighing, and then gently heated before hanging in the solution. The contrast in color between deposited copper and bright platinum is, of course, striking and characteristic. In this way, 0.1 milligramme of copper may be, the author thinks, safely determined; while, for mere qualitative analysis, this method may be employed where the amount is even smaller.

Theory of the Formation of Saline Deposits. In a memoir on the origin of the boracic acid of the Tuscan suffoni, and sundry saline deposits, especially those of Stassfurt, L. Dieulafait (abstract of memoirs in American Journal of Science) lays down the general proposition that "all saline substances existing in mass, or in layers, in sedimentary formations were originally a constituent of a normal sea,' i. e., of a sea of a constitution not essentially different from our present sea. To establish this proposition with regard to the borates, he, in the first place, gives experimental evidence that the water of the Mediterranean contains at least two decigrammes of boracic acid in each cubic metre, and further, that, in evaporating the brine, boracic acid accumulates in the bittern until after the deposition of the carnallite. In the second place, he insists that, in the very characteristic deposits of Stassfurt, the borates are found above the carnallite, as we should expect if these deposits were formed, as assumed, by the drying up of extensive salt lakes. Again, having confirmed the previous statements that the chief salt-beds of the world are found on two geological horizons, the Lias and the middle Tertiary, he gives evidence that, in the Maremma of Tuscany, where the suffoni occur, there is a saliferous basin of the Tertiary period; and he concludes that the suffoni are not properly volcanic vents, but that the surface-water percolating to the salt-beds-heated, it is true, by volcanic agency

determines well-known chemical changes, from which result the peculiar acid-vapors there discharged. But we can only give here the barest outlines of an argument which is worthy of careful study.

M. Dieulafait also contributes in his paper