tion, it now makes little of it. Thirdly, it is not sustained by the analyses of the varieties of foliated pyroxene-diallage, and the wrongly called hypersthene being essentially identical in composition with common augite of eruptive rocks, and the smaragdite with other crystallized hornblende. This is shown in any work giving full lists of analyses of minerals, and is well understood; yet the introduction bere of a few of the analyses may not be superfluous. Nos. 1 to 5, are of diallage and pseudo-hypersthene, and 6 to 8 of augite crystals from Etna and Vesuvius. SiO, Al,0, FeO MnO MgO CaO H2O 1. Florence, Diall... 53.20 2:47 8.67 0.38 14.91 19.09 1.77-100-49, Köhler. 6. Etna, Augite Cryst. 7. Vesuvius, 8. Vesuvius, 66 17.24 15.63 2.13 15:33 1854 1·46 52.34 3.05 8.84 15.58 19 18 0.66 53.60 1.99 8.95 0.28 13:08 21.06 0.86 99.61, Regnault. 99.94, v. Rath. 99.65, Streng. 99.82, v. Rath. The mineralogical and chemical differences are thus too slight to make the distinction of any lithological importance, and this importance can be sustained, if at all, only on geological considerations. The particular rock, in the description of which the character stands prominent, is that called Gabbro in Germany. It is well known that this Italian word was the provincial name originally of common serpentine. Ferber in his "Briefe aus dem Wälschland" (Letters from Italy), written in the years 1771, 1772, and published in 1773, describes so well the rock near Florence, that we cite briefly from him. He first says, in a letter from Florence of Dec. 11, 1771, (in which he gives scientific notes on the minerals and rocks of the region) that the Gabbro of the Italians, occurring in Italy, Tuscany and Genoa, is identical with the serpentine of Saxony. Then, in another, of May 23, 1772, he repeats the statement and describes particularly, and with scientific precision, the gabbro of Mt. Impruneta, near Florence, and mentions the occurrence in it of a talky micaceous mineral which affords, he says, a powder greasy to the touch (the diallage) and also amianthus. He then adds that "in horizontélen Schichten in den Gabbro-Bergen um Impruneta findet sich der sogenannte Granitone, welcher aus weissen Feldspat, der an einigen Stellen Kalchspatartig ist und mit Säuren brauset, etwas grünlichtem silberfarbigen wür flichten Glimmer, und grünlicher Serpentin-Erde, besteht:" a description that distinguishes the gabbro from the granitone. Further, he says, that some of the granitone consists of the "white feldspar in large parallelopipeds and green gabbroearth, without the micaceous mineral." The word Gabbro, as it is now used (and was so first by von Buch in 1810) is applied to the granitone, the associate of the Italian gabbro; but, besides this, to rocks consisting of foliated pyroxene (sometimes called hypersthenite), and cleavable labradorite the idea of foliated standing out prominently; and also to an eruptive diabase-like or doleryte-like rock, in which the augite happens to be foliated. In this last variety, as the analyses show, there is evidently no foundation whatever for separating the rock from other labradorite-augite eruptive rocks. Granitone is the same as euphotide, a rock distributed at intervals along the Alps from Savoy and Isère in France through Piedmont, to the valley of the Saas, north of east of Monte Rosa, and the Graubündten, occurring also in Silesia and on the island of Corsica, and found commonly associated with serpentine. Its chief characteristic is--not its foliated diallage or smaragdite (either of which is usually a mixture of hornblende and pyroxene), but its consisting largely of the compact jade-like material called saussurite; for it would be the same rock, essentially, whether the hornblende and pyroxene were distinctly foliated or not; and, in fact, in part of it the texture is aphanitic, and nothing foliated is distinguishable. Saussurite has a close relation to some of the feldspars in its constituents, it being essentially a soda-lime-alumina silicate; and still, as has long been recognized, it is not a feldspar. This has been rightly sustained by the fact of the high density, which is over 2·9 (2·9 to 3-4), in saussurite, and less than 2.765 in the feldspar group. It is further proved by its occurrence occasionally under the crystalline forms of a triclinic feldspar, but with a fine granular or aphanitic structure; thus having, instead of the cleavage structure belonging to the feldspar, a feature belonging to a pseudomorph. In such cases it was once feldspar; but some change has come over it that has resulted in a molecular transformation, affecting both the crystalline character and the density. Saussurite appears to cover a group of minerals, like feldspar. One kind is between anorthite and zoisite in composition, though differing from both in the soda and magnesia, and from all feldspars in its not having the feldspar-ratio between the silica and soda. A second has the composition of labradorite; and a third the composition nearly of oligoclase. A fourth, from Corsica, analyzed by Boulanger, is a lime-alumina silicate, like anorthite and zoisite. The saussurite group, with density between 29 and 34, runs nearly parallel with the feldspar group. The first is Saussurite, Th. de Saussure having named thus the Lake Geneva variety, after his father, in 1806; the third is Jadeite; and the second may be called, from one of its localities, Genevrite. The following are analyses of the three prominent kinds, and of normal anorthite, labradorite and oligoclase. Specific gravity of 1, 3.227; of 2, 3-3-3-4; of 3, 3·16; of 4, 3·10; of 5, 2·998; of 6, 2.74; of 7, 2·95; of 8, 2·954; of 9, 2·923; of 10, 3-33-3-35; of 11, 332; of anorthite, 2.66-2.763; of labradorite, 2-67-2-76; of oligoclase 2.56-2·72. To No. 9, add erО, 0·51, and to 11, ZnO 0-73. Nos. 10 to 12 are only known worked into ornaments, but the kind may yet be found in the Alps. No. 6 has the specific gravity of labradorite and was therefore that species, a mineral that would be present where the crystallization took place without, or with only partially, the conditions needed to produce saussurite. No. 9 is of the globules of the "Variolite of Durance," a rock associated with euphotide. Boulanger's saussurite, from Corsica, is near zoisite in composition and density (G. 3.18), as stated by T. S. Hunt, who referred all true saussurite to zoisite (confirming his view by his analysis above), and the part near labradorite to that of feldspar. Damour obtained for jadeite the ratio 1: 2: 6. The relation to the feldspar group indicates the occurrence of special geological circumstances, which turned feldspathic material into saussurite. The circumstances that determined the crystallization or metamorphism may have produced, in its incipient stage, soda-lime feldspar; but it ended in making a large part, or the whole, saussurite. Moreover the hornblende has been shown to be in part, at least, pseudomorphous after pyroxene; so that the foliated ingredient bears like evidence in favor of this mode of origin. Consequently saussurite rocks not only differ molecularly from any labradorite or feldspar rock, but are indications of peculiar geological operations on a large scale; and this, connected with other differences, makes it desirable to distinguish such rocks by a special name. The saussurite, and not the foliated mineral, is the chief ingredient on which the distinction rests. Euphotide is therefore a different rock from any consisting of cleavable labradorite and pyroxene or hornblende, both on mineralogical and geological grounds. The foliated condition of the latter constituent is not reason enough for overlooking the more fundamental differences. As the name gabbro has covered rocks of so different kinds, lithology would be freer of ambiguities without it. The true labradorite-and-pyroxene rock of Scandinavia, the Adirondacks, British America and other regions, sometimes called Noryte-the third kind of gabbro-has the chemical and mineralogical constitution of diabase or doleryte. But it differs from these in its granitoid aspect and geological relations, and is of metamorphic origin; and as it is of wide geographical distribution, geology seems to require for it a distinct name, and noryte is an appropriate one. The pyroxene, though generally foliated, is not always so. When, in place of pyroxene, there is true hypersthene, a mineral of different composition and character, as at St. Paul's, Labrador, the rock is then rightly called Hypersthenyte, and this name is so used by Zirkel. 3. Porphyritic Structure.-Porphyry naturally took the position of a species in the mineralogy of the ancients. But it is now well known, and generally admitted, that the porphyritic structure is largly due to conditions attending the former temperature and cooling of the rock-mass, and distinguishes only varieties. But still it is usual to find dioryte divided, for its primary subdivisions, into ordinary dioryte and dioryte-porphyry; diabase into granular diabase and diabase-porphyry or diabase-porphyrite; felsyte into felsyte and felsyte-porphyry; and so on, as if the porphyritic structure were deserving of first prominence in the question of division into varieties, even greater than mineral constitution; and sometimes it is even made the basis of a distinct kind of rock. But, first, this porphyritic feature is only one grade in the crystalline condition, and is of no more value as regards rockdistinctions than other grades. Secondly, it is of far less importance in this respect than any variations in chemical or mineral compositions, such as are made the basis of other varieties. Thirdly, it has often little stability in a rock-formation; for transitions in a dioryte from porphyritic dioryte to nonporphyritic are often found to take place at short intervals, laterally as well as vertically; and so it is with other porphyritic rocks. Within three miles west of New Haven, Connecticut, a labradorite-dioryte undergoes many such transitions in intervals of a few rods, illustrating the little value of the distinction based merely on this condition in the feldspar. Half a dozen miles farther west there is porphyritic granite which graduates in a few yards at some points into porphyritic gneiss (the crystals of orthoclase, two inches long and threefourths of an inch broad) and this last graduates near by into ordinary gneiss; and gradations from porphyrytic to ordinary gneiss are very common in the region. Such facts make it evident that the porphyritic structure is a characteristic of little relative importance; that a porphyritic variety may have rightly a place on a level with other ordinary varieties, but never above one based on variations in composition. The porphyritic structure is an easy character to observe, but this is not an argument in its favor that science can entertain. Such names as felsite-porphyre, amygdaloporphyre, granitoporphyre, melaphyre (this last signifying "black porphyry") and others (abbreviated sometimes to felsophyre, amygdalophyre, granophyre, etc.) have high authority. But they seem to belong rather to books on polished stones than to scientific works on lithology. The occurrence also of the augite of an eruptive rock in distinct crystals, or of quartz in double pyramids, and other similar cases, can have nothing more than a small varietal value. The criterion-crystals or not-is sufficient to distinguish only varieties in mineralogy; and lithology can rightly make no more of it. [To be continued.] ART. XXXIX.-On the Spectrum of the Corona; by W. T. SAMPSON, U. S. N. SO FAR as I can learn from the daily papers, there seems to be a considerable unanimity of opinion as to the nature of the light of the corona. My observations, made at Separation, with Prof. Newcomb, lead me to a somewhat different conclusion from those above referred to; I therefore ask leave to describe my work. Like many other observers of the eclipse I had set myself the task of determining the source of the light of the corona; and this I attempted only so far as to decide by the presence or absence of dark lines in the spectrum of the corona, whether the light was the reflected light of the sun or due to the selfluminous material, or whether it was due to both these combined. For this work I used a direct vision five-prism spectroscope, made by Browning, attached to an equatorially mounted telescope of 3 inch aperture and about five feet focal length. In addition I had prepared to use a hand polariscope consisting of a double-image prism and mica plate. These were the best tools for the work at my command. A few days before the eclipse Mr. Lockyer asked me, if time permitted, to examine with a radial slit whether the 1474 line came close down to the sun's surface or broke off some distance above it. Previous observers were somewhat at variance as to the fact |