the quantity of water in each of these salts, and afterward the percentage of metallic nickel. This last determination was ef fected by first carefully heating the salt in a platinum crucible, employing the ring-burner of Dr. Gibbs so as to apply the heat at the rim of the crucible first, and afterward in successive zones until the bottom of the crucible was reached. The remaining carbon was then burned off in a current of pure oxygen, and the oxide of nickel finally reduced by igniting it in carefully purified hydrogen. The double cyanides of nickel, brucine and strychnine were prepared by double decomposition, the salts being but slightly soluble in cold water. They were then repeatedly recrystallized, and when tested by the spectroscope were found to be absolutely free from potassium. All of these salts crystallized in very pale yellow needles. My analyses of the brucine salt led to the following results: 0-4496 gr. dried at 120° C. lost 0·0258 gr. OH2 =5738 p. c. water. The formula Nig Cy12 (C23H26N2O4), H&+100H2, requires 5.929 p. c. The probable error of the mean percentage of nickel is ±0.008, and the atomic weight of nickel 57.98, with a probable error of 0·089. Six analyses of the strychnine salt were then made: 0-3399 gr. dried at 112° C. lost 0-0178 gr. water = = 5.24 p. c. The formula Ni3Cy12(C21H22N2O2) H ̧+80H,, requires 6 5.45 p. c. The probable error of the mean percentage of nickel is 10013, and the atomic weight 58-038, with a probable error of ±0.119. The mean of all my determinations of the atomic weight of nickel is 58.01. The following table gives all the determina tions made: In conclusion, my thanks are due to Dr. Gibbs for the selection of the subject of my work, and for his advice during the course of my investigation. Cambridge, May, 1871. ART. X.-Note on the Spectrum of the Corona; by Prof. C. A. YOUNG. IN an article upon the Solar Corona, which appeared in the May number of this Journal, I wrote, "very perplexing also is the fact that the faint continuous spectrum, which must be in part produced by this polarized component of the corona's light, shows no discoverable traces of the dark lines of the ordinary sunlight-spectrum. Probably they exist, but are in some way masked so that they are not easily detected." On further reflection, however, I believe the matter is readily explained, and that on the other hand it would have been remarkable if we had been able to bring out the Fraunhofer lines. The truth is that the reflected photospheric sunlight forms only one small fraction of the total coronal radiance, the other constituents of which so far preponderate that it becomes very difficult to detect in the general spectrum the characteristics of this reflected light. The spectrum of the corona is, in all probability, composed of at least four superposed elements. 1st. A continuous spectrum, without lines either bright or dark, due to incandescent dust-that is, to particles of solid or liquid meteoric matter near the sun. For although I am not able to admit with Mr. Proctor that the whole explanation of the corona is involved in the presence of such meteoric particles, yet it cannot be doubted that they are very numerous; and any that may come within 250,000 miles of the solar surface must become incandescent and give such a spectrum as described. 2nd. A true gaseous spectrum of the second order, consisting, like all such spectra, of a more or less bright continuous background with well marked maxima or bright lines. In this case one bright line (1474) certainly exists, and perhaps several. So far as the spectroscopic evidence goes this gas may be simply the vapor of the meteoric dust above alluded to, liberated by the heat of the sun, as when powdered sodium is dropped into an alcohol flame; or it may be disengaged for the instant from the same particles by electric discharges between them, as when the bright lines of metallic vapor appear in the spectrum of the spark produced by an induction coil. But in my previous article I have stated reasons for believing that the gas is of a more permanent character-a solar atmosphere through and in which the meteoric particles move as foreign bodies. 3rd. A true sunlight-spectrum (with its dark lines) formed by photospheric light reflected from the solar atmosphere and meteoric dust. To this reflected sunlight undoubtedly is due most of the polarization, and were it possible to separate the polarized component of the coronal light from the rest, we might perhaps hope to find in it traces of the Fraunhofer lines. It is by no means impossible, however, that the spectroscopic character of reflected light may undergo some change, such as a partial obliteration or degradation of its lines, when the reflecting particles are sufficiently minute-small, that is as compared with the dimensions of a wave of light, so that they do not merely reflect the undulation at their surfaces, but themselves enter into motion bodily. Experiments upon the spectrum of the light emitted by one of the so-called actinic clouds would, perhaps, clear up the subject. 4th. Another component spectrum is due to the light reflected from the particles of our own atmosphere. This is a mixture of the three already named, with the addition of the chromosphere spectrum; for while at the middle of an eclipse the air is wholly shielded from photospheric sunlight, it is of course exposed to illumination from the prominences and upper portions of the chromosphere. This light from the terrestrial atmosphere, like that reflected by particles near the sun, is evidently partially polarized in radial planes. And if there is between us and the moon, at the moment of eclipse, any cloud of cosmical dust, the light reflected by this would come in as a fifth element. It would, however, only dif fer from that reflected by our own atmosphere by including a greater or less modicum of photospheric sunlight. Furthermore, in instruments like those employed by Messrs. Abbay and Pye, the chromosphere spectrum overlies that of the corona, and increases the complication. It would seem, therefore, that only a small percentage of the light which falls upon the slit of the spectroscope during a total eclipse contains the Fraunhofer lines at all, and it ought not to be considered strange that they are not readily observed. In the same article I have stated that the photographs, taken by the American party in Spain, appear to differ essentially from those obtained by Mr. Brothers, in Sicily. This statement was based upon a comparison instituted by Mr. Lockyer, Professor Winlock and myself, between a copy of the American photograph and a drawing* of Mr. Brothers' photograph, which (drawing) he had himself sent to Mr. Lockyer. There was a general and even striking agreement between the two in respect to the position of the gaps' and the distribution of the luminosity, yet there certainly were, as Mr. L. pointed out, very noticeable and important differences, and of a character to suggest that the extensive outside radiance might probably be of a less permanent character than the leucosphere, and of a different origin. But I understand that when photographic copies of Mr. Brothers' and the American negatives are made to a common scale then these differences disappear and the agreement becomes nearly absolute in respect to all essential particulars. If this be so, it certainly bears very strongly in favor of those theories which assign a purely solar origin to the whole phenomenon. Dartmouth College, May 10, 1871. SCIENTIFIC INTELLIGENCE. I. CHEMISTRY AND PHYSICS. 1. Note on Para-sulphobenzoic Acid; by IRA REMSEN, M.D., Ph.D., Assistant in the University laboratory, Tübingen, Germany. -In my former note, (volume i, page 462), I stated that when ordinary sulphobenzoate of potassium is fused with hydrate of potassium, a mixture of oxybenzoïc and para-oxybenzoïc acids is always formed, instead of pure oxybenzoïc acid, as was up to that time supposed. From this fact I concluded that two salts must be contained in ordinary sulphobenzoate of potassium. In order to decide this point, I separated the free acid from the same potassium salt that had been made use of for the first experiment described in the note mentioned. This was accomplished by decomposing the salt with the necessary quantity of sulphuric acid, evaporating to dryness and extracting with alcohol. With the acid mixture obtained I prepared an acid barium salt, by neutralizing with carbonate of barium, dividing the solution into two equal parts, precipitating the barium from the one exactly with sulphuric acid and then mixing the two again. On evaporating this solution, it became filled with beautiful needles on cooling. * I am not sure but we had a photographic copy of Mr. Brothers' drawing instead of the drawing itself; but we did not have a photographic copy of the original negative. No such copies had then been made. These did not possess the least resemblance to the crystals of the known acid barium salt of sulphobenzoïc acid. They were filtered off and recrystallized. Their form was not changed by this process. The salt was now repeatedly recrystallized from water and each time it separated in needles which became longer and more beautiful, the finer they became. It was dried by being allowed to stand for some time over sulphuric acid, and then subjected to analysis. It contained water of crystallization, which was not driven off at a temperature lower than 250-260°. The salt has the formula (C4H10S2010) Ba+3H2O. The analyses gave the following numbers: Found, H2O 8.95 pr. ct. and 9.38 pr. ct. Ba 25.34 pr. ct. and 25.38 pr. ct. Calculated, II,O 9-10 pr. ct. Ba 25:41 pr. ct. The mother liquor from these needles was now evaporated, and in this way a salt of entirely different appearance was obtained. After being recrystallized it formed very regular, beautiful monoclinic crystals, which resembled the known acid sulphobenzoate of barium in every respect. This salt was also analyzed and the same formula found for it as for the needles. The water of crystallization escaped at 200°. (Found, H2O 9:34 pr. ct. Ba 25.56 pr. ct. 25.41 pr. ct. S Although the decided difference in the solubility of the salts and in their crystalline form, which constantly presented itself, made it exceedingly improbable, it was still possible that two different conditions of the same salt were here under observation and not two isomeric salts, particularly as the analyses had shown them to contain the same number of molecules of water of crystallization. There was hence still an experiment necessary to prove the isomerism. The two barium salts were converted into the neutral potassium salts and these melted into hydrate of potassium. The potassium salt that was obtained from the needles, yielded perfectly pure para-oxybenzoic acid as the only product of the reaction. The acid showed all the characteristic signs of para-oxybenzoïc acid. The crystalline form was the same. It contained water of crystallization, which was given off at 100°. Its melting point was exactly 210°. The crystals were very regularly formed and possessed a mother-of-pearl luster. The nature of the salt that crystallizes in needles is thus explained. It is a salt of para-sulphobenzoic acid. Acid para-sulphobenzoate of Barium is difficultly soluble in hot water (much more so than the known corresponding salt of metasulphobenzoïc acid), and almost insoluble in cold water. When pure it crystallizes from a hot solution during the process of filtering. If it be now redissolved and allowed to stand quietly, beautiful flattened needles are found in the solution, which fill the entire vessel from top to bottom. |