mined with certainty (p. 211); but it is possible that further research in the Harz may sooner or later confirm the facts observed in Bohemia, of the sporadic appearance of fishes towards the close of the Silurian period, along with Dalmanites, Bronteus, Acidaspis, and other types characteristic of zone G (p. 212). Moreover, it is possible that the deposits of the Harz form part of the great Paleozoic zone of the north like those of the environs of Hof, and were separated from the Bohemian basin by an ancient barrier of gneiss and other crystalline rocks. In Spain, Sardinia, and France no beds higher than zone E have been observed, unless it is in the department of the Lower Loire, where M. Cailliaud has observed the coexistence of several species characteristic of stage F, with others considered as Devonian.

Crossing the Atlantic, M. Barrande considers the North American representatives of his third division of the Silurian system, and arrives at somewhat different conclusions from those drawn by MM. de Verneuil, D. Sharpe, and Hall. Excluding the Clinton and Medina groups as passage-beds between the two great divisions of the system, he includes the Niagara group and the, whole of the Lower Helderberg series as representing his zone E. Zone F is represented by the Oriskany Sandstone, the remaining members of the upper Helderberg series being regarded as representing generally zones G and H; and he considers that, by means of certain Devonian forms in the Upper Helderberg series, these beds are more intimately connected with the Devonian system of Europe than are those of Bohemia, which contain a smaller number of Devonian types.

Recapitulating his facts, M. Barrande observes that the countries in the neighbourhood of Bohemia, namely, Saxony, Thuringia, and Franconia, do not present a single trace of the last phase of the third fauna, but that the first phase of this fauna is distinctly recognized. The Silurian rocks of these countries appear to belong to the great northern Palæozoic zone, like England, Russia, and the Harz, the faunas of which present us with more marked connexions with the phase of life appertaining to zones G and H.

Norway and Sweden, which belong to the same northern zone, and are intimately related to England by the character of the earlier portions of its Upper Silurian fauna, present none of the phases of the latter part of that period. France, Spain, and Sardinia on the contrary, although forming, like Bohemia, part of the great central Palæozoic zone, and presenting the earlier phase of the third fauna, are equally devoid of the latter portion; but there are beds, considered in France as representing the Lower Devonian system, which contain fossils similar to forms disseminated through zones E, F, and G.

In England, the Isle of Oesel, and the Harz, zones G and H appear to be partially represented. In the Upper Ludlow and passagebeds of England we find six genera of fishes associated with fossils of Silurian types; but the Trilobites Dalmanites and Bronteus, and the Cephalopods with a contracted orifice, characteristic of the higher beds of Bohemia, are absent, as also are the Goniatites and Cardiola retrostriata. In the Isle of Oesel, where the number of fishes amounts to 43 species, among which are Coccosteus and Asterolepis,

there are associated with them numerous fossils characteristic of the earlier period of the third fauna, whilst the Crustaceans and Cephalopods appertaining to the latter part of that fauna are entirely wanting. In the Harz, however, the presence of Ctenacanthus among fishes and of Dalmanites and Bronteus among Trilobites, associated with forms of Devonian type, renders it probable that the limestones of Mägdesprung represent an epoch near those represented by stages G and H (p. 266).

In the United States of America the closest analogy to the fauna of G and H is found in the Upper Helderberg group (especially in the State of New York), which, by its Gasteropods, Acephala, and Brachiopods, presents clearer connexions with the Devonian fauna of Europe than do the similar zones of Bohemia.

With respect to the connexions between these two higher zones and the Devonian system, after passing in review all the evidence to be derived from organic remains, the author arrives at the conclusion that generally they are not more directly united to the Devonian fauna than are the higher zones of any of the other Silurian countries where the third fauna has a less vertical development, and is even reduced to its first phase; for if in Bohemia they contain the genus Goniatites, which is not met with elsewhere, in the former they contain several genera of fishes, and of other types, as Calceola, Grammysia, Pleurodictyum, &c., which occupy a distinguished rank among Devonian fossils, and which are completely wanting in the Bohemian Silurian rocks. And, with reference to specific connexion, the two upper stages in Bohemia contain fewer species identical with Devonian forms, and fewer representative forms, than do the inferior stages E and F. In fact his researches lead him to the apparently paradoxical conclusion that the two highest members of his third fauna present less strong connexions with the Devonian system than do those which immediately precede them.

Taken, therefore, from a palæontological point of view stages G and H preserve completely a Silurian character, and are not united to the Devonian system except by those usual links which announce the approach of a succeeding period. [H. B. H.]

ELEMENTS of CHEMICAL and PHYSICAL GEOLOGY. By Dr. GUSTAV BISCHOF, For. Mem. G.S., &c. Vol. II. Second Edition*.

[Lehrbuch der chemischen und physicalischen Geologie. Von Dr. Gustav Bischof, For. Mem. G.S. Zweite gänzlich umgearbeitete Auflage, Band II. Bonn, 1864.]

In this volume of the new edition of Dr. Bischof's well-known Elements of Chemical and Physical Geology,' the whole of the material of the former edition has been re-arranged, more especially with a view to condensation, but, as a large amount of additional matter has been incorporated, the dimensions of the work are still very great, the

* Volume i. of this edition was noticed in Quart. Journ. Geol. Soc. vol. x. Part II. Miscel. p. 13.

volume in question extending to not less than 952 pages of large octavo size; the bulk would have been still further increased if all the analyses contained in the first edition had been given in full, instead of being merely referred to in foot-notes, as is done in several instances. While in the first volume the general chemical principles involved in the changes of the earth's crust are set forth, the second treats of the salts occurring in the mineral kingdom, taking the widest sense of the term and according to the ordinary chemical classification. The volume is divisible into three parts, the first comprising the Haloid salts, the second the oxy-salts of acids which are easily soluble or miscible in all proportions with water, and the third those of the nearly insoluble silicic acid, or silicates. A fourth division treats of the most important of the minerals which, in the author's opinion, have been produced from the decomposition of silicates. The volume may in fact be regarded as a purely chemical mineralogy, in which, without referring to their external characteristics, the composition, formation, metamorphism, replacement, and decomposition of the nonmetallic minerals is considered, the metallic minerals being reserved for a third volume.

The division into chapters is continued from the first volume, commencing with chapter xvii. and ending with chapter xlii. The first chapter (xvii.) contains general remarks on the alkaline and earthy chlorides, bromides, iodides, and fluorides, and on their presence in crystalline rocks and lavas. Chapter xviii. is devoted to chloride of sodium, under the headings of rock-salt and salt-lakes; the former is shown to be deposited from sea-water by slow evaporation, on account of its great purity, rock-salt being perhaps the nearest approach to a pure chemical compound occurring in nature, some varieties yielding by analysis 100 per cent. of chloride of sodium. The salt from brine-springs, on the other hand, being produced by rapid concentration of the solution, decrepitates when heated, portions of the mother-liquor containing uncrystallizable chlorides being entangled among the crystals. The great difference in the relative proportion of gypsum and common salt existing in sea-water and in saliferous rocks is explained by showing that, although sea-water contains 16.4 times more salt than gypsum, the smaller absolute amount of the latter salt represents 63 per cent. of the total necessary for saturation, while the chloride of sodium only amounts to 7 per It is evident therefore that the solution will be sufficiently concentrated to allow of the deposit of gypsum when 37 per cent. of water has been evaporated, whereas it will be necessary to remove not less than 93 per cent. before any salt can be separated. In this way the great masses of gypsum and anhydrite accompanying the salt-masses of the Alps may be supposed to have been formed, by partial concentration of sea-water which was removed before the salt had time to be deposited. The notices of salt-lakes contain numerous interesting particulars of the composition of the water of the Elton Lake in the Crimea, which is remarkable for the large amount of chloride of magnesium, and the variable amount of common salt, that it contains at different seasons of the year.

cent.

Chapter xix. treats of fluorides, the author noticing the combinations

of fluorine with other elements; fluor spar, its production, association, and pseudomorphic forms, and the more complex minerals cryolite, topaz, and pycnite. It is suggested that the latter may be formed from cryolite, a double fluoride of sodium and aluminium, when the fluoride of sodium is replaced by a silicate of alumina, topaz being produced when the same replacement affects a portion of the fluoride of aluminium.

In chapter xx. the alkaline and earthy carbonates are treated at considerable length, commencing with the degree of solubility of artificial and natural carbonates of lime; then follow the methods by which calcareous pseudomorphs are formed. The carbonates of magnesia, baryta, strontia, iron, and manganese are treated in a similar manner, but the interesting subject of the formation of dolomite is reserved for the next volume. The subject of the composition of the natural carbonates of iron, sparry iron ore, clay-ironstone, black band, &c., is most elaborately treated, no less than fifty-seven analyses of these minerals being given.

Chapter xxi. treats of the different earthy sulphates, gypsum, Epsomite, heavy spar, and celestine, their various methods of occurrence, transformations, &c. The difficulty of finding an origin for the great masses of sulphate of baryta common in the newer mineral deposits, no baryta-salt being recognizable in the composition of crystalline rocks, is adverted to, as are also the recent researches of Max Mitscherlich, who announces the presenceof baryta, varying in quantity from 0·45 to 2·33 per cent., in felspars from the Eifel and S. Gothard. The production of sulphate of baryta by the double decomposition of solutions containing witherite and gypsum is described at length, with numerical illustrations.

Chapter xxii. on phosphates, commences with a notice of the various minerals containing phosphoric acid other than the ores of the heavy metals; the author then treats of Fownes's investigations on the presence of phosphates in crystalline rocks and lavas, and proceeds to consider the solubility of the different substances containing phosphate of lime in water saturated with carbonic acid. This varies very greatly, apatite requiring 393,000 parts of water for complete solution, while the scrapings of the bones of a freshly killed ox dissolve in 4000 parts, and precipitated neutral phosphate of lime requires only 1500 parts. Phosphate of lime is formed when alkaline phosphates are heated with silicate of lime. When phosphate of lime dissolved in carbonated water is brought into contact with alkaline fluorides, fluoride of calcium is deposited and phosphate of soda remains in solution. This reaction is suggested as a probable cause of the increase of fluorine in fossil bones. Phosphate of iron (Vivianite) is noticed at some length, and a description is given of its occurrence in the bones forming the skeleton of a drowned miner, which was found in the Scharley zinc-mine in Silesia, upon reopening the workings after they had been closed for three centuries. Vivianite may be produced by double decomposition from apatite and sulphate or carbonate of iron. Notices of the best known occurrences of phosphates

* Pogg. Ann. vol. cxi. p. 351.

n vegetable and animal structures, as well as in sedimentary rocks, with remarks on the continued circulation of phosphoric acid from the minerals, through plants to animals, and its return through the decomposition of organic remains to the mineral state, close the chapter.

Boracic acid and borates form the subject of chapter xxiii.; the occurrence of boracic acid in the Suffioni of Tuscany associated with sulphate of ammonia and chloride of ammonium is described; and the decomposition of boracite in the limestone from which the hot springs rise is suggested as a means of origin. The formation of tourmaline, &c., is reserved for the chapter on the origin of granite.

The third and most important part of the work, comprising chapters xxiv. to xli., extending to nearly 600 pages, treats of the various silicates found in nature, and the general remarks at the commencement treat of their probable composition and origin. Of the former subject it is remarked that the formulæ in use for the more complex silicates are of little other value than that of expressing the individual views of the chemists by whom they are propounded. Throughout this portion of the work the composition of the minerals is expressed chiefly by their oxygen quotients, that is the fractions obtained by dividing the amount of oxygen in the protoxide and sesquioxide bases taken together by that of the silica. With regard to the origin of crystalline silicates, the author is disposed to consider them as essentially formed by the action of water from amorphous masses; thus he remarks that, as a rule, lavas, especially those of newer date, are not crystalline, although when sufficiently thick they may take years in cooling: this view is especially brought out in the consideration of the origin of leucite, a mineral which occurs most abundantly and in the largest crystals in Vesuvian lavas of pre-historic origin. Even at ordinary temperatures glass becomes crystalline when exposed to atmospheric changes for a long period, as has been shown by Brewster and Zirkel in mediæval glass taken from the windows of Saint Andrew's and Cologne Cathedrals.

The system of arrangement of the descriptive notices of the silicates is based upon the relations of their pseudomorphs to each other, the series being commenced with the zeolites, which never appear except in their own proper forms; felspar is made to take the second place, from the fact that orthoclase has been found in the forms of laumonite and analcime, and therefore may under certain circumstances be derivable from zeolites. The last members of the series are the difficultly decomposable substances: mica, chlorite, steatite, serpentine, &c., which are found as pseudomorphs of almost all the harder silicates. Especial reference is made to the remarkable contrast between the extreme divisibility of mica due to physical structure and its almost absolute stability against chemical changes from the action of the air. This property is assumed to be due to peculiar isomeric conditions of the component silicates. As might be supposed in a substance forming the ultimate product of alteration of a great number of different minerals, micas vary very much in composition, the maxima and minima of the chief components are given as follows:

VOL. XXII.-PART II.

C