If however, we adopt the higher temperature of 626° F. to which there is as yet no well established objection, we may obtain an equal power with much smaller cylinders. The tables show that it is practicable to obtain nearly 9 pounds of available mean pressure under this temperature, when working only against the atmosphere; and this moreover with an economy which, though not absolutely the highest, is considerably higher than any that is attainable under a lower temperature, and nearly two and a half times superior to that of steam of five atmospheres. Now if we compress air into the air chamber to the extent of five atmospheres, and work two cylinders of each set against this pressure, while the third works against the atmosphere alone, we shall require no larger diameter than 45 inches for each of our cylinders to enable us to obtain 1750 horse power, which is equal to that of a first class steamer. If then we increase the diameter to 48 inches, we shall have more than 2000 aggregate horse power, which enables us to allow nearly a sixth part for friction and other drawbacks.* The weight of the cylinders themselves will no longer be an objection. That of the air-chamber, heaters and refrigerators may be a more serious matter; how far it will be so will soon be settled by the experiments of Capt. Ericsson. It may be remarked, however, that the apparatus must be a heavy one, indeed, which will materially outweigh the boilers of the ocean steamers.†

In making these estimates of power, I have not overlooked the manner in which the velocity of the piston is controlled by the increased or diminished resistance, as pointed out by Maj. Barnard in Appleton's Magazine for October. It will be found, upon the principles of estimate adopted by him, that too low rather than too high a velocity of piston is here assumed. I admit the conclusiveness of all that Maj. Barnard has said in regard to the insufficiency of the original Ericsson engines to perform the task

* If again all these cylinders work against the reservoir, 38 inches will be as large a diameter as is necessary, and 40 inches will afford nearly 200 horse power surplus. All the estimates of power made in this paper have been founded on suppositions of pressure to which no serious objection can be taken. But as the power is always directly as the pressure in the air chamber, there is plainly no limit to its increase except the strength of materials. The question of bulk or compactness is also dependent upon the same considerations. With a pressure of 20 atmospheres in the air chamber, a single twenty-inch cylinder (the other suppositions remaining as in the text) would give 300 horse power; and a forty-inch, 1200. Two thirty-six inch cylinders would give 1950 horse power; more than enough to drive the largest oceanliner, after making every deduction for waste and unproductive expenditure of power. Two six-foot cylinders would furnish equal power with but five atmospheres of pressure in the air-chamber. But the cylinders themselves would have to endure a pressure of more than twenty atmospheres. These are the dimensions adopted by Capt. Ericsson for his working cylinders in his new engines; but as the supply-cylinders are less than our supposition, and the compression which the air undergoes in them also less, (as it must be so long as it is proposed to make the regenerators of any use,) the power which these engines will develope must be correspondingly inferior. They are considered and compared with the abandoned engines, in a note published in the last number of this Journal.

imposed upon them. The thoroughness with which he has examined that part of the subject leaves nothing further to be said. But nothing which he or any one else has asserted or proved in regard to those engines in any manner invalidates the truth of the following conclusions, viz:

1. That the elastic force of heated air is a force available for all the purposes for which stationary powers are required.

2. That it is an eminently economical source of power; being in this respect superior to steam, as usually employed, in the ratio of two or three to one.

3. That while it appears to be at present more doubtful to what extent it is applicable, if at all, to the purposes of ocean navigation, its value in that respect remains yet to be experimentally settled.

University of Alabama, Nov. 2, 1853.

ART. XIII-Researches on Globuliferous Rocks; by M. DELESSE, Ingenieur des Mines, Professeur Honoraire de Geologie à la Faculté de Besançon.

[THIS important paper by M. Delesse, is published in the Transactions of the Geological Society of France, and is one among the many contributions of its author to our knowledge of the structure of rocks and their minerals.* We give our readers an abstract of it, presenting the prominent facts and the author's conclusions. The article is illustrated by several elegant plates, only a few figures from which we have copied.-EDS.]

Under the name of GLOBULIFEROUS ROCKS, those rocks are designated which contain certain minerals disseminated through

The following are the titles of some of the papers of M. Delesse on these subjects. 1. Note sur le Chrysotil des Vosges.—Ann. de la Soc. d'Emulation des Vosges, vi, 2nd Cahier, 1847.

2. Recherches sur les Verres provenant de la Fusion des Roches.-Bull. de la Soc. Geol. de France, [2], iv, 1380, 1847.

3. Notice Sur les Caractères de l'Arkose dans les Vosges.-Bib. Univ. de Geneve, March, 1848.

4. Procédé Mécanique pour déterminer la Composition Chemique des Roches.— Bib. Univ. de Geneve, July, 1848.

5. Observations sur la presence d'eau de Combinaison dans les Roches Feldspathiques.-Bull. de la Soc. Geol. de France, [2], vi, 393, 1849.

6. Sur le pouvoir Magnetique des Roches.--Ann. de Ch. et de Phys., 1849, xxv, 194, and Annales des Mines, [4], xiv, 81, 429, and xv; xvi, 323, 1849.

7. Recherches sur le Porphyre Quartzifère.-Bull. de la Soc. Geol. de Fr. [2], vi, 629, 1849.

8. Sur le Porphyre Amygdaloide d'Oberstein.-Ann. des mines, [4], xvi, 511,

1849.

9. Recherches sur l'Euphotide.-Bull. Soc. Geol. de France, [2], vi, 547, 1849. 10. Sur la Constitution Minéralogique et Chemique des Roches des Vosges: Pegmatite avec tourmalines de St. Etienne.-Ann. des Mines, [4], xvi, 1849.

11. Sur la Constitution Minéralogique et Chemique de la Syenite du ballon d'Alsace, Mem. de la Soc. d'Emulation du Doubs. 1847.

12. Sur la Protogine des Alps.-Bull, de la Soc. Geol. de France, [2], vi, 230, 1849. 13. Serpentine des Vosges, Ann. des Mines, [4], xviii, 309, 1850.

These globules are

them more or less thickly in globules.* generally feldspathic, and this memoir has especial reference to those of this kind, occurring in rocks that are rich in silica.

Granites are sometimes globuliferous, as those of Rappawiki, in Finland, which contain globules consisting of orthoclase, surrounded by oligoclase. But these researches relate particularly to porphyritic or compact rocks, especially eurite, pyromeride, trachyte, retinite, perlite, obsidian, and a large variety of porphyries. The pyromerides of Corsica and the Vosges, the porphyries of Esterel, of the country of Bade and Thuringia, the trachytes of Iceland, the retinites of Saxony, and the perlites and obsidians of Hungary, are taken as types of this structure.

The globules vary much in color, being either black, violet, green, brown, yellowish, reddish, gray or white; usually differing a little from the color of the paste. They are commonly harder than feldspar, when undecomposed, owing probably to the excess of silica; but in perlites they are less hard than feldspar, perhaps because the mineral is in a semi-vitreous state, and also, it may be, because opal penetrates them, as stated by Hausmann and Fuchs. The specific gravity is low, viz. 2-3-2-6; 2·3-2-4, in perlite. This is far less than for quartz, which has G. 2·65. Before the blowpipe, they fuse less easily than feldspars, owing to the excess of silica and small proportion of alkalies.

=

0.84-100, Delesse. 1-76-100-01, Ficinus.

8.50-99-85, Klaproth. 0-30-99-16, Erdmann.

74-38 13-78 1.94 119 085 028 3:57 2 63 2:08-101-00, Forchh. 73-00 14:50 1·00 0-10 1·00

1.75

68:33 11:00 400 2·30 8:33 1:30

3:40

The large excess of silica is the remarkable characteristic, and as a general rule in the pyromeride, the composition of the globules is the inverse of that of the enveloping rock.

Structure. The globules have generally a well defined structure. In most of the pyromerides of Corsica and the Vosges,

14. Recherches sur l'Association des Mineraux dans les Roches qui ont un pouvoir magnétique élevé.-Bull, de la Soc. Geol. de France, [2] vii, 108, 1850. 15. Sur la Variolite de la Durance-Ann. des Mines, [4], xvii, 116, 1850.

16. Recherches sur le Porphyre Rouge Antique et sur la Syénite Rose d'Egypte. Bull. Soc. Geol. de France, [2], vii, 484, 524, 1850.

17. Sur le Porphyre de Lessines et de Quenast (Belgique).-Bull. Soc Geol. de France, [2], vii, 310, 1850.

18. Sur la Constitution Minéralogique du Diorite, Kersantites.-Ann. des Mines, [4], xix, 149, 1851.

19. Sur la Constitution Minéralogique de la Calcaire Saccharoide du Gneiss. -Ann. des Mines, xx, 141, (1851.)

This subject has already received some attention--see especially, VON BUCH, Recueil de planches de Petrifications remarquables, fol.; R. C. VON LEONHARD, Charakteristik der Felsarten, p. 52; AL. BRONGNIART, Essai sur les Orbicules siliceux, Ann. Sci. Nat. [1], xxiii, 166; NAUMANN, Lehrbuch der Geognosie; ROTH, Die Kugelform im Mineralreiche und deren einfluss auf die Absonderungestalten der Gesteine.

SECOND SERIES, Vol. XVII, No. 50.-March, 1854.

22

and in the porphyries of l'Esterel and Oppenau, the silicious and feldspathic parts differ in color: and the distinctions of the two are often brought out in decomposition, the silicious part resisting change, while the feldspar is kaolinised. Acids (especially hydrofluoric) develop the structure in a short time.

The globules may be either solid throughout, or they may contain interior cavities: the former I call normal globules: the latter abnormal.

Globules of both kinds are usually spherical or spheroidal, and have either a radiated, concentric, or irregular, structure; these three kinds of structure may occur in the same globule. They detatch themselves readily from the enclosing rock, and present usually an even surface. They undergo alteration less rapidly than the enclosing rock. These globules sometimes contain crystals of quartz or feldspar, which obviously do not concur to the formation of the globules, and which, therefore, are independent crystals.

The excess of silica in the feldspathic paste, is considered an excess of a solvent, as is admitted by M. Delafosse, for different silicates.

Normal globules may either contain quartz or be free from it. Normal globules without quartz.-When 'globules of this kind have a crystalline structure, there are usually distinct cleavages. They are frequently observed in quartziferous, porphyrites, or eurites. Figure 1 represents one of these globules

from the Eurite of Etival (Vosges), a rock consisting of orthoclase, a triclinic feldspar, a little quartz and mica; the globule approaches in form a crystal of orthoclase, and is cleavable and transparent, and it is surrounded by a thin milky or reddish compact zone of a feldspar which is probably triclinic. In certain micaceous eurites of the Vosges (Minettes), the globules are spherical and consist of feldspathic lamellæ, probably of orthoclase, irregularly aggregated, along with some mica, and they have a reddish exterior like the above. When the structure is radiated, there is but a single system of rays, diverging from the centre; the radiation is often very rude. At Oppenau, the feldspar forms but a small part of the globule, it being surrounded by chalcedony. Yet it is evident that the globule has resulted from the tendency of the feldspar to crystallize and to bring under the same influence, the silica.

The globules in obsidian are usually more or less radiated, and sometimes have an exterior coat or layer. They resemble much those found in the crucibles of a glass furnace when slowly cooled. Perlites and retinites, even when apparently compact, consist wholly of globules which may be distinguished when carefully examined; and the structure is very irregularly and sparingly concentric, with some cross fractures (fig. 3).

Normal globules with quartz.-Globules without quartz pass insensible into those with quartz. The latter appear to differ only in that the quartz has been imprisoned within, in consequence of the solidification of a crystalline crust. The structure is either radiated or concentric, and both kinds may occur together. The rays consist of feldspar needles or conoids separated by quartz; and both at the circumference and centre there are often zones of feldspar alternating with quartz.

The structure of the globules of the pyromeride of Corsica is shown in the following magnified view.