in one and the same rock are either all glassy or all fluid. It therefore seems likely to me that in the porphyroidal rocks, the fluid-inclusions are nothing else but original glass-inclusions, out of which the glass-substance has been entirely decomposed and removed by aqueous solutions (p. 196).”

On the other hand, Zirkel* maintains that "these forms certainly do not occur, as Vogelsang considers, 'in great predominance on cleavage-planes.' It may now and then be the case that planes of the inclusions present themselves on the upper surface of the thin section as extraordinarily fine cracks; in such cases, nevertheless, the latter are certainly only secondary; the mineral bearing the inclusions is liable to fracture most easily in that direction in which its continuity is most interrupted, that in which the arrangement of its inclusions occurs;" and again he states: "In all cases we must conclude that the microscopic fluid-particles in different minerals, as well as those included in the constituents of rocks, were enclosed originally with their formation in a mechanical way."

In opposition to the theory of Vogelsang that these liquidinclusions are cavities which have been filled by secondary inJection, Zirkel further calls attention to the absence of com. municating fissures, none being detected under the highest powers; the presence of bubbles in every (?) cavity of a group; the hermetical sealing up of the cavities, so that in experiment a strong heat is required to produce the expulsion of the liquid, with decrepitation; and the chemical nature of the liquids. Therefore, according to the generally accepted opinion on the subject, in all cases rocks have been formed in presence of, or under saturation by, these substances in liquid or gaseous form.

The character and identity of the liquid and gases which occupy these cavities are determined by certain chemical experiments, founded substantially upon their expulsion by ignition in a miniature glass retort, and their subsequent behavior, on introduction into a solution of some proper reagent (e.g. Baryta-water, for the determination of carbonic acid) or into a Geissler-tube for Spectroscopic indications.

Another simple, but important method is commonly employed for the same purpose. It is found that on warming a thin section of rock or mineral containing

*Mikr. Beschaff d. Min. u. Gesteine, p. 47, note.

these fluid inclusions, upon the stage of the microscope, changes invariably occur in the form, size, position, or velocity of motion of any bubble under observation, and in the volume of the liquid in which it floats. In fact, the relative changes in size of this drop at different temperatures, and the complete occupation of the cavity by expansion of the liquid, with a disappearance of the bubble at an observed temperature, sometimes afford an accurate and ready standard for the determination of the coefficient of expansion, etc., of certain liquids found in these cavities, and thus of their identity. For example, the temperature at which liquid carbonic acid assumes this condition, reaches its "critical point" and fills the cavity, is about 30-32° C., while with inclusions of water, or of saline solutions, which are by far more common, the section may be heated up to the point (about 150° C.) at which the hardened balsam, in which it is mounted, begins to soften, without any material change being produced in the relative sizes of the bubble and liquid. Sorby, in his well-known paper,* and after him, J. Clifton Ward, † have not only deduced from the presence or absence of cavities, and from the character of their contents-fluid, glass, or stone-whether the crystals or rocks in which they occur have been formed from aqueous solution, igneous fusion, or both conjoined; but they have availed themselves of the relative volumes of the liquid and bubble in carefully selected cavities to determine approximately the genetic conditions of their formation, temperature, pressure, and rapidity of cooling. This relationship has depended both on the temperature and on the pressure, usually enormous, under which the rocks became solidified. The coefficient of expansion of the liquid being known, and a certain probable temperature being assumed, at which the rocks in question, granites and elvan, were plastic, the amount of pressure was calculated with its equivalent in feet of superincumbent rock. This last amount was always found to be greater by 15,000 to 20,000 feet, than that indicated by the stratigraphical examination of the region, but this excess was attributed to the lateral pressure produced in the folding and crushing of the mass during its upheaval into mountain ranges.

The only purpose in the cursory review just presented has

*Quart. Journ. of Geol. Soc., 1858, XIV., 453.

Idem, 1875, XXXI., 568, and 1876, XXXII., 1.

been to indicate the main facts of a field of investigation rarely familiar to microscopists, as well as the particular views of certain writers which do not seem to present a wholly satisfactory explanation of some of the following facts.

In the counties of New York and Westchester, in this state, a somewhat micaceous and fine-grained, blackish-gray gneiss occurs in considerable abundance, which possesses no physical characteristics of special importance, except the frequent concentration of quartzose aggregates of fibrolite, iron-garnet, and black tourmaline in thin lenticular seams. I have given particular study thus far to the microscopic character of this gneiss only in specimens collected from the vicinity of the town of New Rochelle, in Westchester county. Its thin section here reveals the following constituents :

Quartz predominates in colorless and angular granules of rather uniform size. Its clear material is sometimes traversed by irregular fissures and generally slightly clouded by long, straight and linear groups of inclusions (X 170); under higher power (X500) these are clearly resolved, mostly into cavities of lenticular, angular, elliptical, and circular forms filled with liquid gas, or liquid with a bubble; the bubble in some groups is stationary, in others exhibits a greater or less mobility, from a tremulous vibration to a lively dance in every direction, many liquid inclusions with bubbles in motion being visible at one time.

On focussing at greater depths, the inclusions are found to lie mostly in planes, of which those of neighboring groups are nearly parallel, with shorter branches either connecting neighboring groups, or bifurcating, and often thinning out completely but indistinctly within a quartz field. There is the greatest variety in the numbers and distribution of the cavities within each plane. In size, they may, perhaps, average about 0.00014 mm., but they vary from minute specks, too small for measurement, up to 0.0054 mm., or larger. Their forms show a wide variation, as already described; but, although generally rounded, most of the cavities present one or more plane-faces, a straight side in cross-section, or even a distinct and sharp outline like that of a quartz-crystal, produced by negative or inverted crystallization-a form often observed by Vogelsang and others. Many of these forms, focussed at different depths, are exhibited in the drawing of a single group (Fig. 2)

near the crossing of two planes. In this, some are completely filled with liquid, many contain bubbles, and in the larger cavities which present projecting prongs, the extremities of the latter seem to be occupied by a second liquid. However, on the examination of several liquid-inclusions in this and other groups, by heating upon the stage to a temperature above 40° C., they were found to be little affected and to consist probably of water.

The letters, a, b, c, etc., are attached in the figure to those cavities in which the bubble exhibited spontaneous motion, the most rapid in the smaller cavities; in all other cavities the bubble was quiet. The motion of these bubbles was found to be entirely unaffected when the beam of light from the mirror, either by daylight or from a lamp, was strained from heat-rays by transmission through an alum-cell. In one instance, and only one, a bubble in rapid spontaneous motion was observed to suddenly stop and remain perfectly quiet, but the cause of the stop (possibly some obstacle projecting from the side of the cavity) could not be determined nor could the operation be confirmed by repetition. The longer and more irregular forms of these cavities presented their greater dimensions generally in the plane of the group.

The remaining constituents of the rock consist of a plagioclase-feldspar, in rather rare and small striated grains; lightbrown hornblende, in straight blades with fine cleavages, but irregularly-rounded terminations, which pass through lighter shades of color, into the next mineral; fibrolite, occurring not only in the colorless needles but in blades and in beautiful wisps, knots, and bundles of parallel fibers, often macroscopic, seamed with cross-fractures. Under a higher power (X500) the delicate needles are also found to be sharply defined blades of the same form, with pointed terminations, which, however, are often rounded, or end abruptly in a transverse fracture, penetrating the quartz mostly in parallel bundles, but with many lying obliquely in a very irregular mesh. Many are also dislocated into a series of joints, nearly in position, like those so frequently seen in quartz-enveloped crystals of tourmaline or beryl; and there are also a large number of angular fragments and scales.

The accessory minerals observed are the greenish-brown biotite, whose transverse sections are striated and brownish

yellow; associated and often contiguous scales of colorless muscovite; pinkish grains of garnet, dotted with colorless spherules of quartz; opaque and black granules of magnetite, and brownish-red films of iron-ochre. It was further observed in every thin section (most satisfactorily under a inch objective) that the rock is traversed by very numerous and exceedingly minute fissures, partly in planes which are approximately parallel, at least within the area of the thin section, and partly as branching cracks in an irregular network. The courses wind slightly, but irregularly, showing little or no relationship to the cleavage-planes of the minerals which they may traverse. In crossing the bundles of fibrolite needles, such a fissure is represented by a minute dark line, by interruption of the continuity of that material, with a somewhat jagged border, the fibers or blades of fibrolite being continued on either side, with splintered ends but generally without dislocation or fault. The fissures sometimes appear to be empty in places, or occupied and darkened by minute ochreous particles. More commonly, however, it appears as a vein of laminated structure, consisting of a finely granulated (×925), yellowish material (crushed fibrolite) on each side, next the walls, but in the center of a delicately laminated and colorless substance, apparently quartz, with a very thin and perhaps empty middle suture. The fluid-inclusions or cavities are extremely rare in the vein within a fibrolite-field, and always exactly along the central suture; they occur only close to the entrance to a quartz-field, or enclosed in the occasional quartz-granules within the interstices of the fibers. Some of these veins thin out and disappear within the fibrolite-fields, and near their terminations are crossed by short blades, with fractured ends. The width of the central laminated band-the true width of the original fissure-varies from 0.00007 to 0.00014 mm., and that of the entire vein, including the fragmentary bands, up to about 0.00135 mm.

Within a quartz-field, on the other hand, where the same fissure crosses that material within the interstices of the fibrolite-network, or, more distinctly, where it crosses an adjacent grain of quartz, its width is confined to that of the central band above described, sometimes still continued in the same capillary crevice, or represented interruptedly by short sections of crevice with clear quaitz between, or very often by a line of