By H. SCHUCHT.

Concerning the separations which take place at the positive pole, the composition of the peroxides, and the manner of their determination, relatively little has been done.

If solutions of the salts of lead, thallium, silver, bismuth, nickel, and cobalt are decomposed by the current between platinum electrodes, metal is deposited at the negative, and oxide at the positive electrode. Manganese is precipitated only as peroxide. The formation of peroxide is, of course, effected by the ozone found in the electrolytic oxygen at the positive pole; the oxide existing in solution is brought to a higher degree of oxidation, and is separated out. Its formation may be decreased or entirely prevented by the addition of readily oxidizible bodies, such as organic acids, lactose, glycerine, and preferably by an excess of oxalic acid; but only until the organic matter is transformed into carbonic acid. In this manner Classen separates other metals from manganese in order to prevent the saline solutions from being retained by the peroxide.

With solutions of silver, bismuth, nickel, and cobalt, it is often practicable to prevent the separation of oxide by giving the current a greater resistance—increasing the distance between the electrodes.

The proportion between the quantities of metal and of peroxide deposited is not constant, and even if we disregard the concentration of the solution, the strength of the current and secondary influences (action of nascent hydrogen) is different in acid and in alkaline solutions. In acid solutions much peroxide is formed; in alkaline liquids, little or none. The reason of the difference is that ozone is evolved principally in acid solutions, but appears in small quantities only in alkaline liquids, or under certain circumstances not at all. The quantity of peroxide deposited depends also on the temperature of the saline solution; at ordinary temperatures the author obtained more peroxide—the solution, the time, and the strength of current being equal—than from a heated liquid. The cause is that ozone is destroyed by heat and converted into ordinary oxygen. With the exception of lead and thallium the quantity of metal deposited from an acid solution is always greater than that of the peroxide.

Lead.—Luckow has shown that from acid solutions—no matter what may be the acid—lead is deposited at the anode as a mixture of anhydrous and hydrated peroxide of variable composition. Only very strongly acid solutions let all their lead fall down as peroxide; the precipitation is rapid immediately on closing the circuit, and complete separation is effected only in presence of at least 10 per cent. of free nitric acid. As the current becomes stronger with the increase of free acid, there is deposited upon the first compact layer a new stratum of loosely adhering peroxide.

In presence of small quantities of other metals which are thrown down by the current in the metallic state, such as copper, mercury, etc., peroxide alone is deposited from a solution of lead containing small quantities only of free nitric acid.

The lead peroxide deposited is at first light brown or dark red, and becomes constantly darker and finally taking a velvet-black. As its stratification upon the platinum is unequal, it forms beautifully colored rings.

Experiments show that the quantity of peroxide deposited depends on the nature of the solution and the strength of the current. In case of very feeble currents and slight acidity, its quantity is so small that it does not need to be taken into consideration. If the lead solution is very dilute scarcely any current is observed, lead solutions per se being very bad conductors of electricity.

Faintly acid concentrated lead solutions give loose peroxide along with much spongy metallic lead. Free alkali decreases the separation of peroxide; feebly alkaline solutions, concentrated and dilute, yield relatively much peroxide along with metallic lead, while strongly alkaline solutions deposit no peroxide.

Dried lead peroxide is so sparingly hygroscopic that it may be weighed as such; its weight remains constant upon the balance for a long time. In order to apply the peroxide for quantitative determinations, a large surface must be exposed to action. As positive electrode a platinum capsule is convenient, and a platinum disk as negative pole. The capsule shape is necessary because the peroxide when deposited in large quantities adheres only partially, and falls in part in thin loose scales. It is necessary to siphon off the nitric solution, since, like all peroxides, that of lead is not absolutely insoluble in nitric acid. The methods of Riche and May give results which are always too high, since portions of saline solution are retained by the spongy deposit and can be but very imperfectly removed by washing. This is especially the case in presence of free alkali.

The author has proceeded as follows: The lead peroxide is dried in the capsule, and there is passed over it pure dry gaseous sulphurous acid in a strong current from a rather narrow delivery tube. Lead sulphate is formed with evolution of heat; it is let cool under the exsiccator, and weighed as such. Or he ignites the peroxide along with finely pulverized ammonium sulphite; the mass must have a pure white color. After the conclusion of the reaction it is ignited for about 20 minutes. The results are too high. The proportion of actual lead peroxide in the deposit ranges from 94 to 94.76 per cent. The peroxide precipitated from a nitric solution may, under certain circumstances, be anhydrous. This result is due to the secondary influences at the positive pole, where the free acid gradually withdraws water from the peroxide.

The peroxide thrown down from alkaline solutions retains alkali so obstinately that it cannot be removed by washing; the peroxide plays here the part of an acid. The lead nitrate mechanically inclosed in the peroxide is resolved by ignition into oxide, hyponitric acid, and oxygen; this small proportion of lead oxide does not exert an important influence on the final result. The quantity of matter mechanically inclosed is relatively high, as in the precipitation of much lead peroxide there is relatively more saline matter occluded than when a few centigrammes are deposited. The peroxide incloses also more foreign matter if it is thrown down upon a small surface than if it is deposited in a thin layer over a broad surface. From numerous analyses the author concludes that in presence of much free nitric acid the proportion of water is increased; with free alkali the reverse holds good.

Thallium behaves similarly to lead. From a nitric acid solution it is thrown down, according to the proportion of free acid, either as sesquioxide only or in small quantities as silvery, metallic leaflets; from alkaline solutions it is deposited as sesquioxide and metal, the latter of a lead-gray color. Thallium solutions conduct the electric current badly. Thallium oxide resembles lead peroxide in color; at a strong heat it melts, becomes darker, and is converted into peroxide, in which state it can be weighed.

Silver.—All solutions of silver salts, except the nitrate, and those containing a very large quantity of free nitric acid or nitrates, deposit electrolytically merely metallic silver. In the above mentioned exceptional cases there is formed a small quantity of peroxide which adheres to the anode as a blackish-gray deposit. The greatest quantity of peroxide is obtained on employing a concentrated, strongly acid solution of the nitrate, and a strong current. If the solution is very dilute we obtain no peroxide, or mere traces which disappear again toward the end of the process. The peroxide is deposited at first in small, dark, shining octahedral crystals; subsequently, in an amorphous state. At 110° it evolves oxygen suddenly, and is converted into metallic silver. It dissolves in ammonia with a violent escape of nitrogen. In nitric acid it dissolves without decomposition and with a red color.

The author uses a galvanic current for reducing silver residues, consisting of sulphocyanide. The salt is mixed with sulphuric acid in a roomy platinum capsule, and a fine platinum wire gauze is used as positive electrode.

Bismuth.—The current resolves bismuth solutions into metal and bismutic acid. The latter is deposited at the positive pole, and in thin layers appears of a golden-yellow, but in thick strata is darker, approaching to red. Its formation is very gradual, and in time it disappears again, owing to secondary actions of the current. On ignition it becomes lemon yellow, and transitorily darker, even brown, and passes into the sexquioxide.

Nickel and Cobalt.—On the electrolysis of the ammonical solution the sesquioxide appears at the positive pole. Its formation is prevented by an excess of ammonia. The author never obtains more than 3½ per cent. of the quantity of the metal. The sesquioxides dissolve in ammonia without escape of nitrogen, and are usually anhydrous.

Manganese.—Manganese is the only metal which is precipitated only as peroxide. It is deposited at once on closing the circuit, and is at first brown, then black and shining. Organic acids, ferrous oxide, chromic oxide, ammonium salts, etc., prevent the formation of peroxide and the red color produced by permanganic acid. In very dilute strongly acid nitric solutions there is formed only permanganic acid, which according to Riche is plainly visible in solutions containing 1/1000000 grm. manganese. On electrolyzing a manganiferous solution of copper nitrate, red permanganic acid appeared in a stratum floating above the platinum disk coated with brown peroxide. No manganese peroxide was deposited. The peroxide adheres firmly to the platinum when the proportion of free acid is small, not exceeding 3 per cent., and the current is not too strong. If the action of the current is prolonged after the peroxide is thrown down, it falls off in laminæ. According to Riche, in a nitric solution the manganese is deposited as peroxide, also at the negative pole. This formation is not directly due to the current, but is a precipitate occasioned by the production of ammonia by the reduction of nitric acid. To determine the manganese in peroxide electrolytically precipitated, it is heated to bright redness in the platinum capsule until the weight becomes constant. The results are too high.

Selenium and Tellurium.—Both these bodies are readily and completely reduced by the current either in acid or alkaline solutions. Selenium is thrown down at first of a fine brownish red, which gradually becomes darker. The deposit of tellurium is of a bluish black color. If the current is feeble, the deposit of selenium is moderately compact; that of tellurium is always loose, and it often floats on the liquid. A strong current precipitates both as powders. The positive pole is coated during electrolysis with a film of a dark color in case of selenium, but of a lemon yellow with tellurium. As in case of arsenic and antimony, the hydrogen evolved at the negative pole combines with the reduced substances, forming hydrogen, selenide, or telluride, which remain in part in solution in the liquid. The reduced metal separates out at the anode in a friable condition.—Zeitschrift fur Analytische Chemie, and Chemical News.

A very careful and important determination of the electrochemical equivalent of silver has been made at the observatory of the Physical Institute of Würzbourg, and the results are that an ampere current flowing for a second, or a coulomb of electricity deposits 1.1183 milligrammes of silver or 0.3281 milligramme of copper, and decomposes 0.09328 milligramme of water, a result agreeing closely with that of Lord Rayleigh recently communicated to the Physical Society. An ampere therefore deposits 4.0259 grammes of silver per hour; Kohlrausch's value is 4.0824, a value hitherto accepted universally. This value is so useful in measuring electric currents with accuracy, and free from the disturbances of magnetism, etc., that it is eminently satisfactory to find the German value agree with that of Lord Rayleigh, which will probably be adopted by English electricians.

Herr Hefner-Alteneck has suggested a new standard light for photometric purposes, which promises to be very simple and effective in operation. The light is produced by an open flame of amyl-acetate burning from a wick of cotton fiber which fills a tube of German silver 1 in. long and 316 mils. internal diameter; the external diameter being 324 mils. The flame is 1.58 in. high from top to bottom; and it should be lighted at least ten minutes before using the light for testing. A cylindrical glass chimney surrounds it to ward off air currents. About 2 per cent. of the light is absorbed by the glass. The power of the flame is that of a standard English candle; and experiments have shown that amyl acetate, which besides is not expensive, is the best fuel for steadiness and brilliance. Neither the substitution of commercial amyl-acetate for pure nor the use of a wick of cotton thread for loose cotton fiber alters the illuminating power; but the wick should be trimmed square across the mouth of the tube, for if it project and droop the illuminating power is increased.

[NATURE.]

In a recent number of the Zeitschrift fur Instrumentenkunde (iv., 42-50, February, 1884), Dr. K. Feussner of Karlsruhe has given a detailed description of a polarizing prism lately devised by him, which presents several points of novelty, and for which certain advantages are claimed. The paper also contains an account, although not an exhaustive one, of the various polarizing prisms which have from time to time been constructed by means of different combinations of Iceland spar. The literature of this subject is scattered and somewhat difficult of access, and moreover only a small part of it has hitherto been translated into English; and it would appear therefore that a brief abstract of the paper may not be without service to those among the readers of Nature who may be unacquainted with the original memoirs, or who may not have the necessary references at hand.

Following the order adopted by Dr. Feussner, the subject may be divided into two parts:

I.—OLDER FORMS OF POLARIZING PRISMS.

In comparing the various forms of polarizing prisms, the main points which need attention are—the angular extent of the field of view, the direction of the emergent polarized ray, whether it is shifted to one side of, or remains symmetrical to the long axis of the prism; the proportion which the length of the prism bears to its breadth; and lastly, the position of the terminal faces, whether perpendicular or inclined to the long axis. These requirements are fulfilled in different degrees by the following methods of construction:

Fig. 1., Fig. 2., and Fig. 3.
Figs. 1., 2., and 3.

1. The Nicol Prism (Edin. New Phil. Journal, 1828, vi., 83).—This (Fig. 1), as is well known, is constructed from a rhombohedron of Iceland spar, the length of which must be fully three times as great as the width. The end faces are cut off in such a manner that the angle of 72° which they originally form with the lateral edge of the rhombohedron is reduced to 68°. The prism is then cut in two in a plane perpendicular to the new end surfaces, the section being carried obliquely from one obtuse corner of the prism to the other, in the direction of its length. The surfaces of this section, after having been carefully polished, are cemented together again by means of Canada balsam. A ray of light, on entering the prism, is separated by the double refraction of the calc-spar into an ordinary and an extraordinary ray; the former undergoes total reflection at the layer of balsam at an incidence which allows the extraordinary ray to be transmitted; the latter, therefore, passes through unchanged. This principle of obtaining a single polarized ray by means of total reflection of the other is common to all the forms of prism now to be described.

Dr. Feussner gives a mathematical analysis of the paths taken by the two polarized rays within the Nicol prism, and finds that the emergent extraordinary ray can include an angular field of 29°, but that this extreme value holds good only for rays incident upon that portion of the end surface which is near to the obtuse corner, and that from thence it gradually decreases until the field includes an angle of only about half the previous amount. He finds, moreover, that, although of course the ray emerges parallel to its direction of incidence, yet that the zone of polarized light is shifted to one side of the central line. Also that the great length of the Nicol—3.28 times its breadth—is not only an inconvenience, but owing to the large pieces of spar thus required for its construction, prisms of any but small size become very expensive. To this it may be added that there is a considerable loss of light by reflection from the first surface, owing to its inclined position in regard to the long axis of the prism.

Fig. 4., Fig. 5., and Fig. 6.
Figs. 4., 5., and 6.

It is with the view of obviating these defects that the modifications represented in Figs. 2 to 6 have been devised.

2. The Shortened Nicol Prism.—This arrangement of the Nicol prism is constructed by Dr. Steeg and Reuter of Homburg v.d.H. For the sake of facility of manufacture, the end surfaces are cleavage planes, and the oblique cut, instead of being perpendicular, makes with these an angle of about 84°. By this alteration the prism becomes shorter, and is now only 2.83 times its breadth; but if Canada balsam is still used as the cement, the field will occupy a very unsymmetrical position in regard to the long axis. If balsam of copaiba is made use of, the index of refraction of which is 1.50, a symmetrical field of about 24° will be obtained. A prism of this kind has also been designed by Prof. B. Hasert of Eisenach (Pogg. Ann., cxiii., 189), but its performance appears to be inferior to the above.

3. The Nicol Prism with Perpendicular Ends.—The terminal surfaces in this prism are perpendicular to the long axis, and the sectional cut makes with them an angle of about 75°. The length of the prism is 3.75 times its breadth, and if the cement has an index of refraction of 1.525, the field is symmetrically disposed, and includes an angle of 27°. Prisms of this kind have been manufactured by Dr. Steeg, Mr. C.D. Ahrens, and others.

4. The Foucault Prism (Comptes Rendus, 1857, xlv., 238).—This construction differs from all those hitherto mentioned, in that a film of air is employed between the two cut surfaces as the totally reflecting medium instead of a layer of cement. The two halves of the prism are kept in position, without touching each other, by means of the mounting. The length of the prism is in this way much reduced, and amounts to only 1.528 times its breadth. The end surfaces are cleavage planes, and the sectional cut makes with them an angle of 59°. The field, however, includes not more than about 8°, so that this prism can be used only in the case of nearly parallel rays; and in addition to this the pictures which may be seen through it are to some extent veiled and indistinct, owing to repeated internal reflection.

5. The Hartnack Prism (Ann. de Ch. et de Physique, ser. iv., vii., 181).—This form of prism was devised in 1866 by MM. Hartnack and Prazmowiski; the original memoir is a valuable one; a translation of it, with some additions, has lately been published (Journ. of the R. Microscopical Soc., June, 1883, 428). It is considered by Dr. Feussner to be the most perfect prism capable of being prepared from calc-spar. The ends of the prism are perpendicular to its length; the section carried through it is in a plane perpendicular to the principal axis of the crystal. The cementing medium is linseed oil, the index of refraction of which is 1.485. This form of prism is certainly not so well known in this country as it deserves to be; a very excellent one, supplied to the present writer by Dr. Steeg is of rectangular form throughout, the terminal surfaces are 19 × 15 mm., and the length 41 mm. The lateral shifting of the field is scarcely perceptible, the prism is perfectly colorless and transparent, and its performance is far superior to that of the ordinary Nicol. The field of view afforded by this construction depends upon the cementing substance used, and also upon the inclination of the sectional cut in regard to the end of the prism; it may vary from 20° to 41°. If the utmost extent of field is not required, the prism may be shortened by lessening the angle of the section, at the expense, however, of interfering with the symmetrical disposition of the field.

6. The Glan Prism (Carl's "Repertorium," xvi., 570, and xvii., 195).—This is a modification of the Foucault, and in a similar manner includes a film of air between the sectional surfaces. The end surfaces and also the cut carried through the prism are parallel to the principal axis of the calc-spar. The ends are normal to the length, and the field includes about 8°. This prism is very short, and may indeed be even shorter than it is broad. It is subject to the same defect as that mentioned in the case of the Foucault, although perhaps not quite to the same extent.

II.—THE NEW POLARIZING PRISM.

This prism differs very considerably from the preceding forms, and consists of a thin plate of a doubly refracting crystal cemented between two wedge-shaped pieces of glass, the terminal faces of which are normal to the length. The external form of the prism may thus be similar to the Hartnack, the calc-spar being replaced by glass. The indices of refraction of the glass and of the cementing medium should correspond with the greater index of refraction of the crystal, and the directions of greatest and least elasticity in the latter must stand in a plane perpendicular to the direction of the section. One of the advantages claimed for the new prism is that, it dispenses with the large and valuable pieces of spar hitherto found necessary; a further advantage being that other crystalline substances may be used in this prism instead of calc-spar. The latter advantage, however, occurs only when the difference between the indices of refraction for the ordinary and extraordinary rays in the particular crystal made use of is greater than in calc-spar. When this is the case, the field becomes enlarged, and the length of the prism is reduced.

Fig. 7.
Fig. 7.

The substance which Dr. Feussner has employed as being most suitable for the separating crystal plate is nitrate of soda (natronsalpeter), in which the above-mentioned values are ω = 1.587 and η = 1.336. It crystallizes in similar form to calcite, and in both cases thin plates obtained by cleavage may be used.

As the cementing substance for the nitrate of soda, a mixture of gum dammar with monobromonaphthalene was used, which afforded an index of refraction of 1.58. In the case of thin plates of calcite, a solid cementing substance of sufficiently high refractive power was not available, and a fluid medium was therefore employed. For this purpose the whole prism was inclosed in a short glass tube with airtight ends, which was filled with monobromonaphthalene. In an experimental prism a mixture of balsam of tolu was made use of, giving a cement with an index of refraction of 1.62, but the low refractive power resulted in a very considerable reduction of the field. The extent and disposition of the field may be varied by altering the inclination at which the crystal lamina is inserted (Fig. 7), and thereby reducing the length of the prism, as in the case of the Hartnack.

In order to obviate the effects of reflection from the internal side surfaces if the prism, the wedge-shaped blocks of glass of which it is built up may be made much broader than would otherwise be necessary; the edges of this extra width are cut obliquely and suitably blackened.

The accompanying diagram (Fig. 8) represents a prism of cylindrical external form constructed in this manner, the lower surface being that of the incident light. In this the field amounts to 30°, and the breadth is about double the length.

Fig. 8.
Fig. 8.

Dr. Feussner remarks that a prism similar in some respects to his new arrangement was devised in 1869 by M. Jamin (Comptes Rendus, lxviii., 221), who used a thin plate of calc-spar inclosed in a cell filled with bisulphide of carbon; and also by Dr. Zenker, who replaced the liquid in M. Jamin's construction by wedges of flint glass.

Among others, the carefully considered modifications of the Nicol prism which have recently been devised by Prof. S.P. Thompson (Phil. Mag., November, 1881, 349, and Jour. R. Micros. Soc., August, 1883, 575), and by Mr. R.T. Glazebrook (Phil. Mag., May, 1883, 352), do not appear to have been known to Dr. Feussner.

The following tabular view of different forms of polarizing prisms is taken from the conclusion of Dr. Feussner's paper:

As an analyzing prism of about 6 mm. clear width, and 13.5 mm. long, the new prism is stated by its inventor to be of the most essential service, and it would certainly appear that the arrangement is rather better adapted for small prisms than for those of considerable size. Any means by which a beam of polarized light of large diameter—say 3 to 3½ inches—could be obtained with all the convenience of a Nicol would be a real advance, for spar of sufficient size and purity for such a purpose has become so scarce and therefore so valuable that large prisms are difficult to procure at all. So far as an analyzer is concerned, the experience of the writer of this notice would lead to the opinion that improvements are to be looked for rather in the way of the discovery of an artificial crystal which absorbs one of the polarized rays than by further modifications depending upon total reflection. The researches of Dr. Herapath on iodosulphate of quinine (Phil. Mag., March, 1852, 161, and November, 1853, 346) are in this direction; but crystals of the so-called herapathite require great manipulative skill for their production. If these could be readily obtained of sufficient size, they would be invaluable as analyzers.

This opinion is supported by the existence of an inconvenience which attends every form of analyzing prism. It is frequently, and especially in projecting apparatus, required to be placed at the focus of a system of lenses, so that the rays may cross in the interior of the prism. This is an unfavorable position for a prismatic analyzer, and in the case of a powerful beam of light, such as that from the electric arc, the crossing of the rays within the prism is not unattended with danger to the cementing substance, and to the surfaces in contact with it.

PHILIP R. SLEEMAN.

By F. STOLBA.

Finely ground zircon is quickly rendered soluble if fused with a mixture of potassium borofluoride and potassium carbonate. The author takes two parts of the former to three of the latter, and prepares an intimate, finely divided mixture, which is kept ready for use.

Of this mixture four parts are taken to one of zircon, thoroughly mixed, and melted in a platinum crucible at a red heat. The mass fuses readily, froths at first and gives off bubbles of gas, and flows then quietly, forming a very fluid melt. If the zircon is finely ground, 15 minutes are sufficient for this operation. The loss of weight is 16 per cent., and is not notably increased on prolonged fusion. It corresponds approximately to the weight of the carbonic anhydride present in the potassium carbonate.

The portion insoluble in water is readily dissolved in hydrofluoric acid, and is converted into zircon potassium fluoride. The chief bulk of the zirconium is found in the aqueous solution in the state of double fluorides. The platinum crucible is not in the least attacked during melting. On the contrary, dirty platinum crucibles may be advantageously cleaned by melting in them a little of the above mentioned mixture.

If finely divided zircon is boiled for a long time with caustic lye, it is perceptibly attacked. It is very probable that in this manner zircon might be entirely dissolved under a pressure of 10 atmospheres.

Potassium borofluoride may be readily prepared from cryolite. Crucibles of nickel seem especially well adapted for the fusion of zircon in caustic alkalies.—Ber. Bœhm. Gesell. Wissenschaft; Chem. News.