There are irregular exterior zones of feldspar of different shades, u; needles or conoids of feldspar converging from the zones towards the centre, and others diverging from the centre outward, the two sets much interlaced; the forms and structure of these conoids are very irregular, and sometimes they become isolated globules. Quartz penetrates the whole, and is also to a considerable extent free and glassy (b), filling the spaces or forming zones. Sometimes the quartz is much more abundant at the centre than at the circumference. The quartz is beautifully brought out by means of hydrofluoric acid; but it should be observed that hyaline quartz is much less readily attacked than chalcedony or opaque quartz.

5.

Abnormal globules.-These globules contain cavities, which may be regular or confused, filled or unfilled; and they have been produced either by contraction or expansion. Globules with cavities formed by contraction are common in the trachyte and other rocks of Iceland. One of these enlarged is shown in fig. 5. The cavities are of irregular form and surface. Sometimes they have been filled with quartz, and occasionally with oxyd of manganese, spathic iron, zeolites, barytes, calcite, etc. At Llanberis in Wales, the globules are sometimes a decimeter in diameter, and the cavities contain quartz crystals, and outside of the quartz a layer of ripidolite. Others in retinite contain one or more layers of chalcedony with sometimes an exterior zone of opal and an interior of quartz crystals, or a kind of jasper in layers or bands.

In other globules the cavities are numerous and confused; and often they have been filled with quartz, which may be quite distinct from the feldspar or shade into it. In Corsica, the cavities sometimes contain specular iron with or without quartz; it is implanted in the quartz when this mineral is present.

The independent crystals of globules, occurring in the feldspathic paste constituting them, may be either quartz, orthoclase, a triclinic feldspar, hornblende, mica, specular iron, pyrites; and occasionally also, these form independent globules. In such globules the radiated structure is generally quite confused. Veins or seams of quartz also at times intersect the globules.

The most invariable characteristic of the rocks containing globules, is their holding an excess of silica; and in a given rock they are especially developed where the silica is most prevalent. By this presence of silica, connected in some cases with the penetration of the rock by veins of quartz, the occurrence or the formation of the globules has been determined. Moreover, all the globuliferous rocks are feldspathic; and the feldspar, which

would ordinarily have appeared as crystals, has taken the form of globules. The excess of silica has acted as an impurity in the mass, hindering the formation of regular crystals and leading to radiated and globular concretions. It may have acted also by causing too rapid a solidification for perfect crystallizations of the feldspar. Besides, there was a kind of repulsion between the feldspar and the excessively silicious paste, which was superior to the molecular crystallogenic forces tending to form crystals of the feldspar, and which by acting upon all sides of the agglomerations, reduced them to spheroidal globules.

In the globules of normal form in the pyromeride of Corsica, -which are characterised by exterior feldspathic layers, the outer very quartzose, the inner less so, and a semi-radiated structure within-it appears that the exterior was first consolidated, since these layers are not only very impure, but also support the convergent needles or conoids of feldspar within; the divergent needles from the centre fill up the spaces between the convergent ones, and arise from a solidification nearly cotemporaneous, beginning at the centre, where there is often a zoned feldspathic paste, as a nucleus, which is very siliceous. The solidification of the feldspar appears to have a repelling action on the silica, driving it either to the circumference or the centre. Thus, cooling from the circumference, and cooling from the centre, may occur together, or the former may alone characterize a globule, where its structure is made up of convergent needles alone. When consisting only of divergent needles, or when throughout zoned, it is considered probable that the solidification. took place simultaneously throughout. The silica fills all the interstices between the feldspar needles and zones; it was solidified after the feldspar as in granite. The needles or conoids usually contain a quartz nucleus, so that they are analogous in structure to a globule; the solidification beginning at the surface, the excess of silica was driven to the interior of the conoid.

The abnormal globules, which owe the presence of cavities to contraction, cannot have derived these cavities from contraction after reaching the solid state. The cavities are too large to be thus accounted for by a cause which will explain only the occurrence of fissures. But in the liquid or pasty state, the circumstances are different, especially as a process of desiccation may have been in progress. Rocks of aqueous origin often contain concretions of concentric structure, which contain large cavities that have resulted from a desiccation of the nodule after the exterior was solidified. They may be argillaceous, silicious, calcareous, etc. Among the argillaceous are the septaria of the London clay; of the silicious, there are the brown zoned quartz of Egypt, which contain quartz crystals within, and also the concretions common in different deposits, especially the chalk, as the klappersteine or Achilleum resonans of the chalk of the Isle of Moen.

The calcareous include pisolites, the priapolites of Castres, the kupstein of the Loess, etc. There are also the ferruginous, consisting of argillaceous carbonate of iron, the Etites; also the cupriferous, such as the balls of carbonate of copper of Chessy.

The cavities of these concretions have been formed by contraction while the mass within was in a pasty state. Such cavities after they have been once formed are often filled by infiltrated material, as carbonate of lime, barytes, spathic iron, etc.

In igneous rocks, this process of contraction by desiccation may often be distinguished. In amygdaloids the cavities of the nodules of chalcedony containing crystallized quartz within, are often angular, as if formed by this means. In the Vicentine, the nodules contain angular cavities like those of the abnormal globules above described. Consequently water has performed a part in the formation of amygdaloidal nodules as well as in the formation of rocks of igneous origin. The silica of chalcedonic nodules was originally, as has been suggested, in a gelatinous state; and when owing to the dense condition of it, there was no contraction, the chalcedony entirely filled the cavity, either as a compact or a zoned mass.

But when the silica solution was very fluid, the chalcedony was left in the bottom of the cavity, as occurs at the Giant's Causeway, where the amygdala partly fill the cavities and have a nearly plane upper surface. Generally the solution was less liquid, though still as Breithaupt has observed, sufficiently so to undergo contraction on drying and afford the cavities within that are now filled with quartz crystals and other minerals. The regular forin of the cavity may be owing to the compactness of the enclosing rock, allowing only of very slow drying. The zones in the chalcedony or quartz are a result of the process, showing perhaps intermissions in its progress, or in the molecular attractions in the material present affording the coloring of the layers.

In some cases the abnormal globules have probably been formed through contraction by fusion. It is important to observe that the independent crystals present in these globules, show no trace of the contraction which has been experienced by the abnormal globules; although often thin or delicate, they have not been in any way distorted, or broken. These facts authorize the conclusion that the crystals are subsequent in origin to the formation of the cavities. Hence the cavities must have been formed during the state of partial fusion, before the crystallization of these minerals took place.

From the characters of the cavities and their surfaces, it is inferred that they have probably resulted from the volatilization of water while the globules were still in a liquid state. It might be objected to this view that normal and abnormal globules often occur together. Still the presence of water is altogether proba

ble, and we cannot conceive of any other volatile ingredient being concerned. It is hence altogether probable that these abnormal globules at first contained silica in the state of a hydrosilicate, or a jelly rich in silica; the water was at first retained by the pressure or the heat causing the spheroidal state; and afterwards it gradually escaped, and thus occasioned the contraction and the formation of the cavities.

In abnormal globules having cavities formed by expansion, these cavities have beyond doubt been produced by a disengagement of volatile substances, which has taken place while the rocks were still fluid. When the volatile substances were simply gaseous, they formed no deposit in the cellules; but when they carried along other substances, cooling would lead to a crystallization of different minerals on the walls of the cellules. These minerals are those composing the rocks, especially feldspar, quartz and mica: and in fact it has been found at Sangerhausen that feldspar may be deposited by sublimation in furnaces.*

In some cases, there has been an expansion by gaseous substances, producing an enveloping cellule, and afterwards a contraction forming a feldspathic globule at the centre of the cavity. In other cases, the expansion has taken place within the globule while the paste was still soft, producing a multitude of small celThese cellules have been afterwards filled by chalcedony. Filling of the cavities.-The filling of the cavities has taken place either by secretion, infiltration or by both combined. In the filling of septaria, there may be a secretion of carbonate of lime or iron along the walls of the fissures, and also an infiltration filling up the rest of the cavities, the two processes producing deposits of different kinds or colors. In the filling of a mineral vein also, both processes have often operated, secretion introducing a part of the minerals, and infiltration another portion.

The infiltration of the silica into cavities, may have been either slow or rapid; the slow producing crystals, the more rapid forming chalcedony. Both processes, secretion and infiltration, may at different times have acted. Agates of the melaphyres have resulted from the penetration of the cavities with gelatinous silica, which is easily dissolved in hot water, when this water is under pressure and is charged with carbonic acid.

The same causes account for the threads or veins of quartz which intersect the globules.

The feldspar of the globules may be either orthoclase or a triclinic feldspar. The abnormal globules rather than the normal, are developed in globuliferous rocks, which are rich in soda. When orthoclase and a soda feldspar occur together, the latter is in

* Scacchi has recognized in a recent paper, (Rendiconto della R. Acad., Napoli, 1852,) that mica, quartz, garnet, hornblende, pyroxene, sodalite, nepheline and other silicates may form by sublimation.

the smaller crystals; and this inferior tendency to crystallization thus indicated, may be the reason for the formation in such rocks of abnormal globules which are less crystalline than normal globules.

In conclusion it may be observed that the theory here proposed is universal in its application to all globules or concretions, both those of sedimentary and igneous rocks.

One deduction of interest deserves to be noted. It is that as the vapor of water and volatile substances have acted an important part in the formation of globules, the theories for explaining the origin of granite and feldspathic rocks, ought necessarily to include this agency; for it follows that the feldspars, including orthoclase, may be formed even in the presence of water.

ART. XIV.-Examination of some Deep Soundings from the Atlantic Ocean; by Prof. J. W. BAILEY, West Point, N. Y.

In an account of a microscopical examination of soundings made by the U. S. Coast Survey near the Atlantic coast of the United States* I made known that the soundings along the coast, from the depth of 51 fathoms S. E. of Montauk Point, to 90 fathoms S. E. of Cape Henlopen were chiefly made up of vast amounts of Foraminiferous shells, rivalling in abundance the deposits of analogous fossil species which I had proved to compose immense beds under the city of Charleston, S. C.

The facts were also mentioned that none of the species found in the soundings belong to the littoral genera of the group Agathistegnes of D'Orbigny (Plicatilia, Ehr.) and that they also differed from those found in the tertiary deposits of Maryland and Virginia. These facts were confirmed and extended by the observations of F. de Pourtales in his Report to Prof. A. D. Bache, on the distribution of Foraminiferæ on the coast of New Jersey as shown by the off-shore soundings of the U. S. Coast Survey.† In this paper Mr. Pourtales states that "the greatest depth from which specimens had been examined is two hundred and sixty-seven fathoms, and there the Globigerina are still living in immense numbers." He adds that the region of Globigerina extends to a depth not known.

I am indebted to that zealous cultivator of science, Lt. Maury of the National Observatory, for an opportunity to examine the deep sea soundings made by means of Brookes's lead on board the U. S. Dolphin by Lt. Berryman. These soundings proved to be of great interest and furnished results which have an important bearing upon Geology and Physical Geography.

*See Smithsonian Contributions to Knowledge, vol. ii, Art. 3.

See Proceedings of American Association for the Advancement of Science, 1850, p. 84.