by fusion, both by the study of natural and artificial products, with a few apparent exceptions, which receive a special explanation.

By a judicious combination of substances and temperatures, the authors succeeded in obtaining eleven distinct mineral associations, almost exactly reproducing, even in the minuest details of structure, as many natural rock types.

These are as follows: 1°. Augite (oligoclase) andesite, 2°. Augite (labrador) andesite, 3°. Augite (anorthite) andesite (all produced by single fusion at temperature No. 3); time three days. 4°. Basalt. Two successive stages of fusion were necessary to produce this rock. Temperature No. 2 produced in forty-eight hours numerous crystals of olivine embedded in a glassy matrix, which was altered into a crystalline mass of labradorite and augite microliths by being again subjected for an equal length of time to temperature No. 4. 5°. Nephelinite was produced in forty-eight hours at temperature No. 4. 6°. Leucitite was obtained after three days' fusion at temperature No. 2. 7°. Leucititephrite produced by double fusion exactly like basalt. 8°. Lherzolite, 9°. Meteorites free from felspar, and 10°. Felspathic meteorites, though quite successful so far as the mineral associations were concerned, showed certain variations from the natural products in their structure. No synthesis was perhaps so interesting as that of 11°. Diabase, with the so-called ophitic' structure. This structure consists, as is well known, of irregular masses of pyroxene filling the spaces between the lath-shaped crystals of plagioclase. It was found to be impossible to reproduce this structure with oligoclase or labradorite, on account of their comparatively low fusingpoint. By means of a double fusion with anorthite, it was, however, successfully accomplished.

Scarcely less interesting than these positive results are the conclusions derived from the authors' negative experiments. It was found impossible to obtain the acid rocks, i.e., those containing either quartz, albite, orthoclase, muscovite, biotite, or amphibole, by purely igneous fusion. These minerals either produced an amorphous mass, or passed into other combinations giving rise to species already obtained; e.g., hornblende, when melted, crystallized as pyroxene. Thus the very important conclusion is reached, that the acid rocks owe their origin to some other agency than simple fusion.

Under the head of the synthesis of minerals, the authors' experiments in fusing mixtures

of felspars are worthy of special notice as being directly opposed to the now generally accepted theory of Tschermak, that the triclinic felspars form an isomorphous series. Fouqué and Lévy found it impossible to obtain crystals of intermediate members, as only well-defined microliths of either oligoclase, labradorite, or anorthite, appeared, varying in their relative proportions with the mixtures fused. Also of especial interest are their artificial production of felspars with lead, barium, and strontian as bases.

THE GEOLOGY OF NATAL. Natal. Department of mines. Report upon the coalfields of Klip River, Weenan, Umvoti, and Victoria counties, together with tabulated statement of results obtained from a series of trials of colonial coal upon the Natal government railways. By F. W. NORTH. London, Harrison, pr., 1881. 1, 66 p., (49) pl., etc. fo.

THIS report contains two maps, showing the distribution of the coal-fields of the colony of Natal, and a description of 72 sections occurring in them, 70 of which are illustrated by diagrams. There are also two horizontal sections given, -one from Buffalo River to the Drakensberg Mountains, and the other from Buffalo River to Elands Laagte.

Mr. North estimates the actual area of the Natal coal-field, where he has found workable coal-seams at the surface, at about 1,100 □ miles, situated entirely in Klip River county. To this he adds 250 miles for the region between the Ingagani River and the Drakensberg Mountains, which he considers the coal measures underlie. The workable seams vary from 4 to 10 feet in thickness, and are of several qualities. Assuming an average thickness of 4 feet, and allowing a deduction of 50 per cent for faults, worthless coal, and barren ground, he estimates the whole at 2,073,000,000 tons, divided into,

brown iron ore, and yielded on analysis 63.51 per cent of metallic iron.

Accompanying the report is a "Horizontal geological section on the main road from Durban to Van Reenen's pass, by Dr. P. C. Sutherland." This, in so far as it covers the same ground, differs considerably from that published by C. L. Griesbach in 1871. The Table Mountain sandstone, referred by Griesbach to the carboniferous, is by Sutherland considered Silurian. The mesozoic eruptive rocks. are joined together under the name of basaltic, and are apparently represented as dikes, and not as interstratified flows of melaphyr, amygdaloid, and aphanitic diorite, as by Griesbach.

Mr. North gives the following geological order of succession in the rocks of Natal:

1. Basaltic trap rocks, often penetrating between stratified rocks or shales of the coalmeasures, and forming horizontal beds.

2. Triassic horizontal coal-measures, containing coal-seams correlating with the Stormberg coal-field of Cape Colony.

3. Pietermaritzburg shales, probably corresponding with the upper Karroo beds of Cape Colony.

4. Conglomerate or bowlder clays, in all probability the Dwyka conglomerate of Cape Colony.

5. Sandstones, horizontal and massive, of the Inanda location, Table Mountains, and Bothas Hill, etc., probably of Silurian age. 6. Primary rocks, granite, gneiss, marble,

etc.

Mr. North seems to have overlooked the cretaceous series, from the lower greensand up to the white chalk described by C. L. Griesbach in south-eastern Natal; and no evidence is given for assigning the Table Mountain sandstone to the Silurian instead of the carboniferous in fact, no notice whatever is taken of Mr. Griesbach's able work on the geology of Natal.

At the Insiswa Mountains, in the Amaponda territory, the line of demarcation between a vast eruption of igneous rock and the triassic contains various ores of copper containing traces of gold. Mr. Griesbach also mentions the occurrence of copper ores along the line of the eruption of melaphyrs. We have here, in another portion of the world, another instance of the occurrence of cupriferous traps in the trias.

The bowlder clay consists of a bluish gray base, so fine that its constituents are not resolvable except under high magnifying power, and then no crystals are disclosed. It appears

to be a very fine indurated mud, containing bowlders, pebbles, angular fragments, and grains of a great variety of rocks varying in size from masses weighing over 5 tons to pieces smaller than a pea. In mechanical composition it greatly resembles the great Scandinavian drift. It stretches for hundreds of miles, and has been found 1,200 feet thick. Some of the larger angular bowlders seem to have been brought from a distance of at least 70 miles. It seems difficult to account for such a formation otherwise than by glacial action at the close of the dyassic period.

THE BIBLIOGRAPHY OF ANGLING. Bibliotheca piscatoria. A catalogue of books on angling, the fisheries, and fish-culture, with bibliographical notes and an appendix of citations touching on angling and fishing from old English authors." By T. WESTWOOD and T. SATCHELL. London, Satchell, 1883. 397 p. 8°.

THE possibilities of the future in the formation of libraries on special subjects, at present rates and ratio of increase in book-making, are brought forward in a striking manner by examination of a list like that before us. Here is a work devoted to angling, fisheries, and fishculture, in which 2,148 distinct publications are registered under 3,158 entries, inclusive of new editions and reprints. Angling occupies 245, fisheries 83, and fish-culture 23 pages. Roughly estimated, nearly ten per cent of the publications, including reprints, etc., have appeared since 1870. Fish-culture alone claims an increase of nearly one-third in the same time. It is hardly to be expected that a work of this character should be entirely exhaustive or complete. The authors deserve great credit for the nearness of their approximation to completeness, for the amount and quality of information given, and for general accuracy.

An example or two will indicate respects in which the book may be improved in future editions.

"Gesner (Conrad). Aqvatilivm animantivm

nomina Germanica et Anglica, serie literarum digesta, authore Conrado Gesnero. [1530?] 8°. Appended to an edition. P. Ovidii Nasonis Halieuticon, etc.' Tigvri apud Gesneros fratres, pp. vi+ 280, and extending from page 12 to 280..."

I. Scholia et emendationes in Halieuticon P. Ouidii Nasonis. [pp. 1-11.] [Second title.] P. Ovidii Nasonis Halievticon liber.

II. Aqvatilivm Animantium Enumeratio iuxta Plinium, emendata et explicata serie literarum. [pp. 12-92.]

[Running titles.] Catalogvs Aqvatilivm, and Divisio Aqvatilivm.

III. Eorvndem Nomenclator Germanicus longe copiosissimus. Et alia quaedam ad Piscium historiam pertinentia. [pp. 93-280.]

[Running title.] Teütsche namen der Fische vnd Wasserthieren.

Tigvri apud Andream Gesnerum F. [Date of Prefaces 1556.]

In consequence of the foregoing, after Ovidius Naso (Publius), "Halieuticon: hoc 'est, de piscibus libellus, mute quam ante hac emendatior et scholiis illustratus. . * per Conradum Gesnerum. Tiguri apud Gesneros fratres [1530] 8°" should give place to the

following: Part I. of Gesner's De piscibys et Aqvatilibvs, "Scholia et emendationes in Halieuticon P. Ovidii Nasonis." "P. Ovidii Nasonis Halieuticon liber." pp. 1-11. Tigvri apud Andream Gesnerum F. [1556.]

The date for the first Frankfort edition of Aldrovandi is 1623 instead of 1629; and Gronow gives that of the second as 1640 instead of 1645. That given by the latter as Venice, 1616, is omitted. Three editions of Aelian (1556, 1611, and 1616), given by some authorities, do not appear. Future revision of the work will probably introduce the names of such works as those of Schomburgk's Fishes of British Guiana (1852), and Spix and Agassiz' Fishes of Brazil (1829), both of which give information on angling. The latter figures on plates A to G the various methods of capture in use among the natives.

Our authors have given us a work of great importance to all interested in the subjects of which it treats.

ASTRONOMY.

Mass of a planet from observation of two satellites. M. Struve recommends measurement of the positive angle and distance of a satellite from another satellite, and not from the primary planet. A series of such measurements on satellites of Jupiter has been begun at Pulkova. The observations occupy one-third the time, and are considered two or three times as accurate as those by direct reference to the centre of the planet. They are free, moreover, from the unknown constant errors inseparable from the latter, an advantage which Prof. A. Hall, in this paper, considers cheaply purchased at the price of greater difficulties in computation. He shows, that, while the solution of six normal equations requires seventy-seven auxiliary quantities, that of twelve (the elements of both orbits being involved by the new method) requires four hundred and fortytwo, and therefore nearly six times the labor. But these twelve equations give the period and mean distance of each satellite, and hence two values of the planet's mass. Mr. W. B. Taylor objected to such special designations as 'peri-Saturnian,' apo-Jovian,' for the apsides of satellites' orbits when general names were needed. He suggested 'peri-apsis' and 'apo-apsis.' (Phil. soc. Wash., math. sect.; meeting April 26.) [919 Periodicity of auroras. Professor Sophus Tromholt has discussed the observations of auroras made by Prof. S. Kleinschmidt at Gotthaab, in Greenland, from 1865 to 1880, together with other observations in northern latitudes, and finds that for polar regions the maximum of auroras responds with the minimum sun-spot period, the reverse of what has been noted in temperate zones. The yearly maximum is at the winter solstice, while, in lower latitudes, maxima occur at the equinoxes.

cor

Report of work of the Royal observatory, Cape of Good Hope. In the report for 1882, Dr. Gill states that the observations for the difference of longitude between the observatory and Aden are completed. The great comet was observed on every clear night from Sept. 7, and photographs were obtained on six nights. The heliometer measures for the parallax_ of certain southern stars are nearly concluded. In connection with observations in the northern hemisphere, Victoria and Sappho have been observed for determining the solar parallax by Galle's method. Time of contact at the transit of Venus was noted by six observers, and heliometer measures were made during the transit. not., March, 1883.) M. MCN.

MATHEMATICS.

(Monthly [921

Infinitesimals. Mr. M. H. Doolittle looks on infinitesimals, differentials, and zero as having the same denotation, but different connotations. He proposes, in cases where the value of a function becomes indeterminate, to call that value which is continuous with those for preceding and succeeding values of the variable the serial value. The differential coefficent, in this view, is the serial value of the ratio of two increments when those increments become zero.-(Phil. soc. Wash., math. sect.; meeting May 9.) [922

PHYSICS.

Electricity.

On secondary batteries. - Professor Barker gives a brief history of secondary batteries from the discovery of electrolytic polarization by Gautherot, in 1801, to the invention of the Faure cell, together with the results of his own experiments upon cells of this latest form.

In charging his series of thirty-four cells by means of a Gramme machine, he used, in order to prevent discharge by a current backward through the machine when the electro-motive force of the latter fell, a cut-out,' in which an electro-magnet, through which the current flows, forces the end of a metal bar against a spring, pressing it down, and thus keeping the circuit closed while the current flows in the desired direction. When the current begins to fail, the reaction of the spring opens the circuit.

Using this cut-out, Prof. Barker found that the secondary battery could be employed with great advantage in steadying the current furnished to a series of Edison lamps by a Gramme machine driven by a gas-engine. For this purpose he connected the Gramme and the battery as if for charging, the cutout being in the circuit, and connected, also, the poles of the battery with the lamps. The electromotive force of the machine was made very nearly equal to that of the battery, so that, just after each explosion in the gas-engine, the machine prevailing sent a current through the lamps, and also a small current through the battery, slightly charging it; but, before the next explosion occurred, the electromotive force of the machine had fallen to such a point that the battery now sent a current to the lamps. It is stated, that, although the engine gave only one explosion in four strokes, the pulsations in the light entirely disappeared when the above arrangement was adopted.

Prof. Barker states that his experiments entirely confirm those of Gladstone and Tribe as to the formation of lead sulphate when a secondary cell remains in open circuit. In several cases the acid of the cells disappeared entirely in this way, and lead sulphate formed the entire coating on both plates. On attempting to re-charge such a cell, the resistance was found to be very high, and torrents of gas were evolved from both plates. After a time the resistance fell to its normal value, and the waste of gases ceased, though not till a considerable quantity of energy had been lost.

It appears, moreover, that in the cells employed, although they were intended to be all of like dimensions and construction, there was great difference of storing-capacity and of resistance. After an hour's use on the lamp-circuit, different cells gave on a tangent galvanometer deflections varying from 87° to 0°. When the discharge was continued for a long time, so as nearly to exhaust the battery, it was found that many of the cells were reversed, so as to be now opposing the action of the others. "In place of continuing uniform as a single cell, the electro-motive force of a series of cells begins to fall when about half the charge which it ought to be capable of yielding has been drawn from it." In the Planté cell the local action is far less than in the Faure, the lead peroxide in the former being very much harder, so that not a trace of the sulphate was found in such a cell after six months of frequent use. Prof. Barker appears, therefore, to consider the Planté cell more promising than the Faure, in spite of the much greater time required to form it. (Proc. Amer. assoc.; Montreal meeting, 1882.) E. H. H.

[923

CHEMISTRY.

(General, physical, and inorganic.) Ammonio-argentic iodide. By digesting argentic iodide with a solution of ammonia, A. Lougi obtained the compound NH,AgI. — (Gaz. chim. ital., 1883, 86.) C. F. M. [924

CI

CI

CI

Bleaching-powder and analogous bodies. — The constitution of this substance is again reviewed by Lunge and Naef. In 1882 Kraut objected to the formula, Cl- CaOCl, first proposed by Odling for the dry salt, and apparently confirmed, in 1880, by Lunge and Schaeppi. Kraut's objections were twofold. In the first place, he asserted that all the chlorine in bleaching-powder was expelled by a mixture of carbonic dioxide and hypochlorous acid, and, secondly, that bleaching-powder was analogous to the lithium salt (LiCl + LiOCI). In answering the first objection, Lunge and Naef affirm that Kraut must have started with a basic calcic chloride, which, with hypochlorous acid, gave, first, bleaching-powder, from which carbonic dioxide set free the chlorine, CaOH + HOCI = H2O+CaoC - CaC+ CO2 = CaCO3+ 2C1. They further assert that CaCl, may be decomposed by hypochlorous acid (CaCl2 + Cl2O = CaOCl2 + Cl2). Concerning the second point urged by Kraut, Lunge and Naef find that eighty-eight per cent of lithic hydrate is converted into the basic chloride, while, according to Kraut, fifty per cent only should enter into the reaction, if it is analogous to bleachingpowder. Chlorine is not eliminated from the lithium salt by carbonic dioxide at ordinary temperatures. At higher temperatures the chlorate is formed, and oxygen evolved. The strontium salt corresponding to bleaching-powder, when treated with carbonic dioxide, behaves in a manner strictly analogous to the calcium salt. The authors regard these facts as sufficient to establish the formula, Cl- CaOCl. ― (Berichte deutsch. chem. gesellsch., xvi. 84.) C. F. M.

[925

Action of certain vegetable acids upon lead and tin. — Mr. F. P. Hall tried the action of acetic, tartaric, and citric acids upon lead, tin, alloys of these metals, and upon cans that had been used to preserve fruit. In a solution of approximately the same strength as common vinegar, these acids exerted a much greater corrosive action upon tin than upon lead, whether acting upon the metals separately or in the form of alloys. Both metals were dissolved freely, especially from the cans. The lead probably came from the solder, since it was not detected in the tin of which the cans were made. In the composition of tin foils, every variation was found between samples that were free from lead and those which contained a very high percentage of this metal. (Amer. chem. journ., iv. 440.) C. F. M. [926

(Analytical.)

Direct estimation of chlorine in presence of bromine or iodine, and of bromine in presence of iodine. According to the observations of G. Vortmann, metallic chlorides are not affected when boiled with the peroxide of lead or of manganese and dilute acetic acid, and only with difficulty by the concentrated acid. Bromides are decomposed by plumbic, but not by manganic peroxide, while iodides are readily decomposed by either. To determine chlorine in presence of bromine, the latter may be expelled by evaporating the solution to dryness with plumbic peroxide and dilute acetic acid. Iodine may be expelled from a mixture of a chloride and an iodide by either plumbic or manganic peroxide and acetic acid. Manganic peroxide is also used to decompose an

iodide in presence of a bromide. In a mixture of the three haloid salts, both bromine and iodine may be removed by plumbic peroxide; or first the iodine by manganic peroxide, and then the bromine by plumbic peroxide. These methods fail to give accurate results when the relative percentage of chlorine is small. (Sitzungsber. kais. akad. Wien, lxxxvi. 244.) C. F. M. [927 AGRICULTURE.

--

By-products from rice. The chief by-products of the preparation of rice for market are 'douse,' or bran, rice-flour,' and 'polish.' The bran consists of the hull, or pericarp, with a portion of the outer proteine-bearing layer of the true seed adhering to it. The rice-flour is produced by pounding the grain, freed from the hull, in wooden mortars, to complete the removal of the testa and proteinebearing layer of the seed. It consists of the latter mixed with more or less of the starchy interior portion of the seed. The rice then passes under stiff brushes, which remove the last traces of the outer layer, and more or less starch. The refuse from this process is the rice polish. Analyses of these materials indicate that they are valuable feeding-stuffs, and show them to be decidedly rich in fat and proteine.(Rep. N. C. exp. stat., 1882, 87.) H. P. A.

[928 Analyses of cotton-seed. The following analyses of the hulls and kernels of cotton-seed were made at the North Carolina agricultural experiment-station.

The Atacama (Bolivia) meteorite. This supposed meteorite has been regarded as a pallasite or syssiderite (SCIENCE, p. 41), according as the classification of G. Rose or Daubrée is followed, closely allied to the Siberian form found by Pallas. Dr. S. Meunier dissents from this opinion, although holding that both are specimens of concretionary veins (SCIENCE, p. 18), - - a view for which the present writer is unable to see any basis. A chemical and mineralogical examination showed, according to Meunier, that the non-metallic part had the following composition:

N. ir. extract Fat

The Rastenberg granite. This rock, microscopically studied by Koller, is a porphyritic granite composed of quartz, orthoclase, plagioclase, biotite, and hornblende. These form a medium crystalline mass in which large orthoclase crystals are embedded. Dihexahedral quartz, which is usually present in such rocks, was absent from this. The orthoclase was found to belong to the microperthite variety. While the large crystals were orthoclase, the smaller ones were mostly plagioclase, lying between oligoclase and albite, or, according to Tschermak's theory, between Ab An and Ab, Ang. The absorption and pleochroism of the hornblende were not strong. The colors were, for a, light yellowish-brown; c, clear grassgreen; b, dark brownish-yellow; while a < c < b. — (Min. petrog. mitth., v. 215.) M. E. W. [930

The rocks of the Wechsels. In a paper relating to the lithology of the Wechselgebirge, by Böhm, the rocks are classified as, 1°. Micaceous rocks, divided into albite-gneiss, granulitic-albite-gneiss, micaschist, epidotic-mica-schist, and quartzite; 2°. Chloritic rocks, into chlorite-gneiss and chlorite-schist; 3°. Hornblendie rocks, into diorite-schist and hornblende-epidote-schist. Descriptions of the microscopic characters are given. Rutile and titanite were

Pyrrhotite Olivine

-(Comptes rendus, xcv. 1384.) M. E. W.

The Mocs meteorite. - Professor A. Koch has continued his papers on the fall of meteoric stones in the vicinity of Mocs in the Siebenbürgen, Feb. 3, 1882. He reports from this fall 912 pieces, weighing collectively 174,113 gr. A chemical analysis, made by his brother, Frank Koch, showed that their average composition was as follows:

-

Fe 0.