SCIENTIFIC INTELLIGENCE.

I. PHYSICS AND CHEMISTRY.

1. Spectral analysis of the protuberances observed during a total eclipse of the sun.-The total eclipse of the 18th of August, 1868, attracted an even more than usual degree of attention. It was observed at Wah-Tonne on the peninsula of Malacca by Rayet; at Guntoor by Janssen; at Beejapoor by Capt. Haig, R. E.; at Barram Point on the Island of Borneo by Hennessy; and at Jamkhandi by Lieuts. Herschel and Campbell. The observations made by means of the spectroscope were particularly interesting and important, and agree upon the whole remarkably well. Rayet was provided with a silvered glass reflecting telescope, the reflector twenty centimeters in diameter, and a direct vision spectroscope with three prisms. When the image of a long protuberance on the eastern border of the sun was brought upon the slit of the spectroscope, a spectrum was seen with nine brilliant lines; of these, four corresponded respectively to B, D, E and F; two corresponded to lines in b; two to lines in the group G, and there was one unknown line. The western portion of the sun's margin gave the same lines except that only one violet line was visible. With respect to these observations we may remark that according to the other observers Rayet must have mistaken C for B, and also that the line which he identifies with D is really more refrangible, being separated from it by an interval too large to admit of a doubt as to the want of coincidence. In his letter to the Académie des Sciences Janssen gives no details of the spectral lines observed by him, but communicates the important result that the spectra of the protuberances may be seen at all times, so that it is not necessary to wait for an eclipse. His conclusion is that the protuberances are immense gaseous masses, chiefly composed of hydrogen. Though the merit of having first observed this most important fact belongs undoubtedly to Janssen, it appears that in 1866 Mr. Norman Lockyer suggested that the spectra of the protuberances might be seen by examining the margin of the sun's disc, but did not succeed in observing these spectra until the 20th of October of the present year, or two months after the observation had been made by Janssen. On the 20th of October Mr. Lockyer observed the spectrum of a protuberance, and found it to consist of three lines, one exactly corresponding with C, one very nearly with F, and one eight or nine degrees of Kirchhoff's scale more refrangible than D. There was no ray at B or at b; the region near G was not examined. Observations recently made in Cambridge by Prof. Winlock with the 15-inch equatorial and with a spectroscope with two flint glass prisms of 60°, give the same results. Mr. Lockyer also found that the three bright rays extended to a small distance upon the surface of the sun, and that their lengths were different, the red being the shortest. In a very

interesting and suggestive paper "On the physical constitution of the Sun and Stars," communicated to the Royal Society, May 15th, 1867, Mr. G. J. Stoney remarks that he has several times thought that he observed a bright line somewhat more refrangible than D in examining the sun with a spectroscope composed of two flint glass prisms, as well as two other lines, one between 1025.5 and 1027.7 of Kirchhoff's scale, and another between the two lines D. During the summer of the present year Mr. Stoney saw several faint bright lines in other parts of the spectrum, one coinciding with or very near Kirchhoff's copper line of wave length 522.3. There can be no doubt that the line seen near D was one of those noted by Mr. Lockyer.-Rayet in Comptes Rendus, lxvii, 759; Janssen in Comptes Rendus, lxvii, 838; Lockyer in Comptes Rendus, lxvii, 904; Stoney in Proceedings of Royal Society, xvii, p. 1.

W. G.

2. On the magnetism of chemical combinations.—WIEDEMANN, to whom we owe most of our precise knowledge of the magnetism of compounds, has published a new series of observations of much interest. These all confirm the law already announced by the author that in salts of similar constitution containing the same metal, the magnetism referred to the equivalent of the compound, or the atomic magnetism, is sensibly constant. The author's new results may be stated as follows:

(1.) Observations made by the method used in his former investigations show that the same relations exist for the haloid and oxysalts of cerium, didymium and oxyd of copper (CuO). Taking the unit previously established, we have for the atomic magnetism of aqueous solutions of the salts above mentioned the following

(2.) We obtain approximately the same values for the solid salts, especially when they contain water of crystallization. For anhydrous salts the atomic magnetism is in general a little less than for the hydrated salts. This diminution is particularly sensible for the salts of copper and nickel.

(3.) Experiments prove that two diamagnetic elements--as for example, copper and bromine-may give a magnetic compound.

(4.) The author has examined the magnetism after double decomposition of two saline solutions of magnetism M, and M2 known

before mixture. After reaction the magnetism is sensibly equal to the sum M,+M2. In the greater number of cases the formation of a precipitate exerts no disturbing influence.

(5.) The atomic magnetism of hydrated oxyds is a little higher or lower than that of the corresponding salts.

(6.) The magnetism of precipitated hydrate of sesquioxyd of iron increases rapidly, beginning at the moment of precipitation. The author attributes this fact to the colloidal state of the hydrate at the first moment of precipitation. The differences presented by the atomic magnetism of acetate of sesquioxyd of iron doubtless arise from the same cause. These differences are not remarked in the solution of hydrate of sesquioxyd of chromium, in caustic potash, and in that of oxyd of nickel in ammonia.

(7.) The magnetism of the anhydrous oxyds of the magnetic metals is notably less than that of the saline combinations of these oxyds.

(8.) With the exception of sulphate of manganese, the magnetism of the sulphates of the magnetic metals is very feeble.

(9.) The magnetism of the cyanids of nickel and cobalt disappears almost completely when the cyanids are dissolved in cyanid of potassium, which would not happen if this solution be considered as a double salt.-Comptes Rendus, lxvii, 833, from Monatsbericht der Königl. Preuss. Akad. for July, 1868.

W. G.

3. On a new series of chemical reactions produced by light.TYNDALL has communicated to the Royal Society the results of experiments made by subjecting the vapors of volatile liquids to the action of concentrated solar or electric light. The author's paper is too full of detail to admit of an abstract, but we shall give a few of the more salient results. A tube 2.8 feet long and 2.5 inches internal diameter is closed at both ends by glass plates. It may be connected with an air pump and with a series of tubes used for the purification of air. A number of test tubes were converted into Wolfe's bottles by means of corks and tubes. Each test tube was filled partly with the liquid to be examined and introduced into the path of the purified air. When the experimental tube was exhausted and the air then allowed to bubble through the liquid, a mixture of air and vapor entered the experimental tube together, and were then submitted to the action of light. At one end of the experimental tube was placed an electric lamp transmitting an intense beam of light through the tube parallel to its axis. When the vapor of amylic nitrite was allowed to enter the tube in the dark and the beam of light was then sent through the tube, the tube appeared for an instant optically empty; then a sudden shower of liquid spherules was precipitated on the beam. On repeating this experiment with a condensed beam of light forming a cone eight inches long, the cone which was at first invisible flashed suddenly like a luminous spear. The rapidity of the condensing action diminished with the density of the light. The same effects were produced when oxygen or hydrogen were AM. JOUR. SCI.-SECOND SERIES, VOL. XLVII, No. 139.-JAN., 1869.

employed as carriers; when the head of the beam was sifted out through a plate of alum, or when the beam was used without sifting. That the amylic nitrite undergoes decomposition is proved by the formation of brown fumes of nitrous acid. Sunlight produces similar effects. The author proves in the next place that the decomposition is effected by the more refrangible rays of light, and that liquid amylic nitrite is most potent in arresting the rays which affect its vapor. This seems to show that the absorption takes place in the atoms and not in the molecules. The author anticipates wide, if not entire, generality for the fact that a liquid and its vapor absorb the same rays. When the tube is filled with a rare and well mixed vapor, the electric light develops a blue color, which may be pure and deep or milky according to the intensity of the light. The author connects this result with those of Brücke on the colors of the sky. Various other liquids were tried with success. In many cases the condensed vapors formed extremely beautiful and regularly shaped clouds, the particles rotating around the axis of the tube, or round other axes. The most beautiful forms appear to have been those produced by iodhydric acid.— Proc. of the Royal Society, xvii, p. 92.

W. G.

4. On the carbonaceous matter of meteorites.-The analyses of Wöhler and Cloez have shown that certain meteorites contain substances composed of carbon, oxygen and hydrogen, and resembling the last residues of organic substances of terrestrial origin. Berthelot has applied to the carbonaceous matter of the meteorite of Orgueil the method of hydrogenation employed by him with so much success, and has succeeded in producing a notable quantity of hydrocarbons of the formenic series (CanH2n+2) comparable with the oils of petroleum. This result forms a new analogy between the carbonaceous matter of meteorites and substances of organic origin found upon the surface of the earth.-Comptes Rendus, lxvii, 849.

W. G.

Prof. Brush has called my attention to Mr. Henry's paper, with which I was not acquainted at the time of the publication of my own process for the estimation of manganese as pyrophosphate. That full justice may be done, Mr. Henry's paper is by my request republished in full.-W. GIBBS.

5. On the separation of Nickel and Cobalt from Manganese; by T. H. HENRY, Esq., F.R.S.-The methods given by the best authorities for the separation of nickel and cobalt from manganese are either inconvenient or inaccurate. The only method which affords exact results is that of Ebelmen, in which the sulphids formed at a high temperature are acted upon by dilute hydrochloric acid; but it is both tedious and unpleasant to pass sulphuretted hydrogen over the oxyds contained in a porcelain tray in a porcelain tube at a red heat; and the modification suggested by Wöhler, of converting the oxyds into sulphids by fusing them with sulphur and carbonate of soda, still furnishes the sulphid of nickel in such state that it is slightly acted upon by very dilute hydrochloric acid.

I have obtained very accurate results by a process differing altogether from those described, but as simple as any.

When chlorid of ammonium and ammonia are added to a warm solution of sulphate of manganese or chlorid of manganese, and afterward phosphoric acid till the precipitation ceases, the whole of the manganese is precipitated; and after filtration no precipitate or even cloudiness is produced by the addition of sulphid of ammonium to the solution. The salt formed is, according to Otto, NHO 2MnO, PO,+2HO, and after ignition 2MnO+PO. When a solution of chlorid or sulphate of nickel is treated in a similar manner, no precipitate occurs, even on standing a few days in a vessel lightly covered, when sufficient chlorid of ammonium and free ammonia are present.

The following example will show the mode by which I ope

rate.

12.6 grs. of pure sulphate of manganese, MnOSO3+4HO, were gently ignited, and weighed 8.73 grs. =411 MnO. 6 grs. of oxyd of nickel, containing a trace of silica, were ignited and weighed then 563 grs.; these were dissolved together in hydrochloric acid and water, and the solution diluted; it was made acid with hydrochloric acid, phosphoric acid was added, and the whole heated until nearly boiling: when ammonia was added in excess, a white bulky precipitate was produced which rapidly contracted in volume and became crystalline; after standing twelve hours it was filtered and washed with a solution of chlorid of ammonium and ammonia (the chlorid of ammonium is absolutely necessary). The precipitate was perfectly white, and remained so after ignition, when it weighed 8.67 grs. 4.33 MnO; it was decomposed and examined, it did not contain a trace of nickel.

The ammoniacal solution was treated by sulphuretted hydrogen, and the sulphid converted into oxyd by roasting with a little carbonate of ammonia. It weighed 5.57 grs.

I probably drove off a little sulphuric acid on igniting the sulphate of manganese. I have obtained more accurate results with the manganese since.

In operating with cobalt and manganese in the same manner, I obtained a slight excess of manganese, and the salt was slightly pink, but on repeating the operation the separation was complete. I took 3.22 grs. of oxyd of manganese, and obtained 3.15 grs. free from cobalt.-Phil. Mag., IV, xvi, 197.