189 G e n e r a l a n d P h y s i c a l Chemistry. On some Points connected with the Chemical Constituents of the Solar System. By J. H. G L A D S T O N E (Phil. Nus. [5], iv, 379-385) .-The formation of the solar system by thc condensation of a nebulous mass made up of many different chemical elements, will have resulted in a distribution of those elements dependent on the two followina considepations :- (1.) During the process of cooling the least volatile constituents will condense first and sink towards the centre of gravity, while thc rest will arrange themselves inore or less in the order of their volatility. (2.) As mas pointed out by Mr. C. J. Stoneg, in an atmosphere de- creasing in temperature from within outwards, the lightest molecules will be the farthest from the centre of gravity.The most volatile elements, and those of least vapour-deiisity, will therefore be the outermost in a hot nebulous mass. This distribution is seen to occur in the sun, where hydrogen forms the bulk of the outer atmosphere, mixed with small quantities of sodium and magnesium, while the vapoui* of iron is found only in a lower stratum of the sun’s atmosphere, and platinum has not been detected at all. According to the nebular theory, the planets wcwe formed by the separation of some of the outer portion of the original mass; we should therefore expect them to contain a preponderance of the more volatile elements, and those of least vapour-density ; and this is in- deed the case. Of thecnori-metallic elements, those which are plentiful have an average vapour-density of 19.8, while those wliich are com- paratively rare have an average vapour-density of 63 : grouping the metals into four classes-plentiful, common, rare, and very rare, we find that the average vapour-densities of each class are, 37.8, 104.5, 106.7, and 122.9 respectively.The meteoric stones which fall to the earth from interplanetary spaces show this preponderance of the lighter elements still more strikingly. There are, however, in both cases many exceptions to thc rule, for in- Etance, glucinum and lithium are both rare, while lead is w r y common ; such exceptions may be cxplained by supposing these elements to have been combined with others forming compounds of g1-catcr or less volatility than themselves ; thus carbon is very difficult to volatilise, but its compounds with oxygen, hydrogen, &c., are gases at the ordi- nary temperature.The heads of comets emit light giving band spectra which are usually referred to carbon ; the volatilised carbon of the electric lamp was also found to give these band spectra ; in cornets, however, the carbon is probably combined with oxygen or hydrogen. The Influence of Temperature upon the Coefficients of Refraction of the Natural Sulphates of Barium, Strontium, and Lead. By A. ARZRUNI (Jahrb. f. J&., 1877, 52t;--527).--The author considered it desirable to ascertain whether the coefficients of F. D. €3. VOL. XXXlV. 2’190 ABSTRACT8 OF CHEMICAL PAPERS. refraction remained constant or varied with the temperature a t which the determination was made.Three isomorphous sulphates, viz., barium salpliate, strontium sulphate, and lead sulphate were selected for the investiption ; firstly, because Descloiseaux observed in them a great variation of the angle of the optical axis with the temperature, from which it might justly be inferred that any change in the co- efficient of refraction could easily be determined ; and, secondly, it was desirable to ascertain if isomorphous cornpounds with analogous optical characteristics still remained analogous to an increase of temperature. F o r these observations the author prepared prisnis out of a barytes crystal from Dufton, a cmlestine crystal from Lake Erie, and an anglesite crystal from Monte Poni, the results of the investiga- tions being briefly as follows, vie.:-(1.) The principal coefficients of refraction of the above-mentioned isomorphous sulphntes differ from each other with the temperature, but all of them decrease with an increase of temperature. (2.) The decrease in the coefficient of re- fraction of the three sulphates is an analogous one, and can he ex- pressed thus, viz., Further, y approaches the two others, whilst u withdraws from 6. (3.) With anglesite the refraction is inversely as the temperaftme, whilst the dispersion increases for dif- ferent colours. (4.) The directions of the maximum, medium, and minimum expansicjn of the three compounds by heat do not staid in any relationship to the values of the directions of optical elasticity in them, or to the alteration of the velocity of light in these three direc- tions.C. A. B. > u > 6. On the Law of Absorption and its Employment in Quantita- tive Spectrum Analysis. By G. Govh (Conyt. r m r l . , lxxxv, 1046- 1049).-The author at first alludes to the phenomena of absorption, and compares tlie increase and dilatation o€ absorption-bands ‘‘ due to thickening in the absorbing medium,” to the increase in the bright lines in the spectra of incandescent gases caused by increase in the temperatiwe aEd pressure. He says tliat it is not possible to determine the absorbing power of a body, unless the coefficients of absorption are known f o r all wave-lengths which can be studied. He thinks, however, tlhatj the absorbing power of bodies may be determined, either by direct comparison of the curves of chroniatic absorption, or by measuring tlie intensity of the light along the whole length of the spectrum.To use the first of these methods, lie employs the ab- sorbent in tlie shape of a prism, and by placing one of its plane faces against the slit, with the centre angle touching one end of it, he ob- tains a gradually increasing thickness of the absorbing medium throughout the whole leiigth of the slit, any deviation heing ncu- tidised by a prism of the smallest absorbing power. When white light is passed through the arrangement described above, the spectrum shows more or less wavy shadows, representing to the eye the law according to which tlie coefficient of absorption of the medium varies with the wave-length of the incident light. Should the slit be divided longitudinally into two equal parts, i t is easy t o compare the two chromatic absorption-spectra produced.By using solar light the absorption-curves may be compared with E’raunhof er’s lines. TheGENERAL AND PHYSICAL CHEMISTRY. 1'31 author concludes by saying that in cases requiring great accuracy &is method is not' applicable. J. M. T. Theory of the Action of certain Organic Substances in increasing the Sensitiveness of Silver Haloids. By M. C. L E A (Amer. J. of Xci. [3], xiv, 96-99).-Explanations of this action of certain organic substances have been offered by Poitevin and by H. Vogel, who suggest that' it' is due to their af€inity for the halogen. The author, however, .points out that all the organic suhstances pos- sessing the property in question are reducing agents, and argues that.i t is their affinity for oxygen which, aiding as it does the affinity of tlie halogen for hydrogen, determines the decomposition of the silver-salt. That this is so is provcd by the fact that when pyrognllol is acided to recently precipitated silver iodide, and the mixture exposed t o sun- light, it exhibits a distinctly acid reaction at tlie end of 15 minutes. Had the first-mentioned explanation been the true one, an iodo-substi- tution-product would have been formed, but no acid. Substances which have a great affinity for iodine, as potassinm carbonate and starch, do not increase the sensitiveness of the silver haloid. These facts support tlie views of the author, as does also the process of alkaline development, tlie alkali, 11y ncutralising the acid produced, assisting the development of the image.Y. D, B. A Battery in which the Carbon Electrode is the one At- tacked. By P. JABLOCHKOFP (Conzpt. rend., lxxxv, 1052-1053).- The electricity produced by electro-magnetic machines is due t o the combustion of carbon. The author has attempted to produce electricity by direct action on carbon. As carbon is not attacked by liquids atj ordinary temperatures, he has constructed a hot liquid electro-chemical battery. For this purpose lie uses either potassium or sodium nitrate in which ordinary coke is used as one electrode, and for thc other platinum-iron, or any metal not attacked by the liquid in lwesence of carbon. By the addition of various metallic salts the electro- motive force may he modified, the nietals being deposited on the unattacked electrode.The electromotive force of the battery the author states to vary between 2 and 3 uiiits, while the Runscn battery gives a maximum of 1.8, and Gernet 2.1. To start the action of the battery, a piece of iiicandescent coke is placed in contact with the crushed nitrate, heat) is pyoduced, and the action commences. A large quantity of carbon dioxide and other gases are evolved, which the author proposes to use as a motive power. A description of the appa- ratus is given. J. 31. T. The Movements of Electrified Mercury. By HE R M AS N HE R W I G (8w.n. Ph?ys. Chem. [Z], i, 73--95).-Earlier experiments on the capillary depression of electrified mercury (this Journal, 1877, i, 677) had rendered it probable that the surface of the glass tube in which the mercury was contained was attacked. I n the experiments here detailed the mercury contained in a capillary tube vas electrified by connecting it with one pole of a Holtz machine, while the other pf!192 ABSTRACTS OF CHEMICAL PAPERS.pole was connected with the earth, sparks being meanwhile allowed to pass between the poles which were not too far apart. When the mercury was connected with the positive pole, i t rorc gradually in the tube when the machine was set in motion, forming, a t the highest point to which it reached, and where the surface remained for some time, a dirty ring, which was not removed by repeatedly pouring dean quicksilver through the tube ; this ring held up the thread of mercury when it was no longer electrified.When, on the other hand, the mercury was connected with the negative pole of the machine, it did not rise so high in the tube, no ring was formed unless the electrification was continued for a long time, and on its ceasing the quicksilver sank immediately to its former level ; this latter fact clearly shows that the cohesion of the quicksilver is diminished by electrifica- tion. Drops of mercury placed upon a horizontal glms plate, and posi- tively electrified, rapidly formed a ring of dirty mercury on the plate : when, however, both the plate and the metal hid been carefully dried, these rings were only obtained with great difficulty. The quicksilver is therefore oxidised if water he present. That the diminution of the rapillary depression is not caused by such oxidation alonc is, liowever, shown by the fact that when tm-o similar tubes were filled, the one with hot, dry quicksilver, the other with the same substance inten- tionally moistened, both tubes being closed at the top, the S~TIIC phe- nomena were observed in each on electrification.Furtlicr, when the space above the quicksilver -as filled with dry hydrogen, the same effects were produced as with air, though not so rapidly, although all chance of oxidation was removed. When the mercury of a barometer is positively electrified in thc same manner, the discharge from thc upper surface being facilitated by connecting the outer surface of the glass with the ground, it ap- pears to boil violently, and glowing particles are projected against the sides of the tube ; the effect is much inferior with negative elcctricity.The diflerent action of the two electricities is explained by sup- posing that where a break occurs in a series of conductors, the nega- tive electricity flows more easily, while the positive collects on the surface till a far greater tension has been attained, thus occasioning, on at length passing off, a much greater disturbance. When the mercury is connected with the positive pole the entire 1-acuous space is filled with bluish-green light, which, examined with the spectroscope, is shown t o be mercury light ; close to the mercnry meiiiscus a yellowish-green fluorescent light is also observed which gives an almost continuous spectrum ; when connected with the nega- tive pole the bluish-green light is less developed, while the fluorescent light fills the upper part of the vacuum.The above liypothesis cx- plains these facts also, since the negative electricity passing more easily from the quicksilver surface, the charge would accumulate a t the upper portion of the tube where its presence would be indicated hy the fluorescent light, whereas the positive would be found close to the surface of the mercury. The same explanation applies to the movement of threads of mer- cury in horizontal capillary tubes : it was found, for instance, thatGENERAL AND PHYSICAL CHEMISTRY. 193 when a platinum-wire, connected with the negative pole of the machine, was placed near one end of the thread, the merciiry rapidly approached the wire, whereas i&en it was connected with the positive pole the mercury was almost always repelled.0 bservations made with tubes containing threads of mercury bounded a t each end by water, were attended with like results. The results o€ the experiments already made are summed up as follows :- (1.) Powerful charges of electricity diminish the cohesion of mer- cury more than the adhesion between mercury and glass, thus lessen- ing the capillary depression in glass tubes. (2.) When positive electricity a t high tension escapes from the sur- face of mercury in glass vessels, tlie surface being exposed to the air, oxidation takes place, more especially when moisture is present. Nega- tive electricity has a reducing action. (3.) The passage of electricity (particularly positive) at high tension from mercury to glass occasions decomposition of the glass.(4.) The three facts above mentioned are the united cause of the observed diminution of the capillary depression. ( 5 . ) Negative electricity escapes from highly charged conductors at a less tension than positive. (6.) When surfaces of mercury are charged with electricity, the formation of mercury-vapour is greatly facilitated. Positive electricity has by far the greatest effect. (7.) Mercury-vapour is relatively an excellent conductor of elec- trici ty . F. I). B. On the Specific Heat of Vapours and its Variations with the Temperature. By E. WIEDEMANN (Am. Phys. C h n . [2], ii, 1'35-21 7) .-In Regnault's researches upon specific heats of vapours, he determined the specific heats at high temperatures by finding the qiiantities of heat given up in each case in condensing the gas from two different high temperatures to the same lower temperature, the difference between the two quantities giving the specific heat between the two high temperatures.He found that the specific heat et a t a temperature t, might be represented by the formula c, = c, + gat, when c, is the specific heat a t 0" and 2u the alteration of specific heat for one degree of temperature. His method is objectionable in this : that though the amount of heat given up in each case may be large and capable of determination with only a small proportiorlate error, nevertheless a small error may bear a, large proportion to the dif- ference between two determinations, and so considerably affect the result.Wiedemann avoids this difficulty by using an arrangement in which he can obtain the vapours a t low pressures and therefore at lower temperatures, and he observes directly the heat given off in cooling a vapour from a higher to a lower temperature. He has investigated thus the values of c, and u in Regnault's formula for chloroform, ethyl bromide, benzene, acetone, acetic ether, and ethyl oxide. His resulting specific heats differ from those of Begnault by from 3 to 5 per cent., but agree much more closely among them-194 ABSTRACTS OF CHEMICAL PAPERS. selves, the variations from Regnault's results being probably clue to i 111 purities . He finds in general that the greater the specific heat of a 1iqu:d thfl greater is that of its vapoin. The coefficient a is o€ tlic same order of magnitude for the liquid and its vapour, and in some cases nrarly the same for the two, but it varies greatly for dif-ferent vapours.J. H. P. The Internal Condition and Latent Heat of Vapours.. By P. C. PUSCHL (Chem. Centr., 1877, p. 318).-Bg a metliod incle- pendent of the second law of thermodynamics, the author deduces the general equation which holds for saturated vapours. Hc further shows that, in the cycle of operations upon a mixture of liquid and rapour, which consists (I), in allowing the same to expand at a con- stant temperature; ( 2 ) ' heating a t the rolumc attained and kept constant; (3), allowing it to contract to its initial volume a t this increased temperature; and (4), allowing it to cool to the init'ial temperature while maintaining its volinme constant at t,his point,-the work expended is not the equivalent, but is in excess of the heat ob- tained.There has therefore been a gain in internaJ work, the qnnntitg of which may be determined in the case of water and its vapour from Regnault's experimental data. The values of the forces which deter- mine the volume of water vapour, under the external pressure, may easily be determined for temperatures between 0" and 200". Respect- ing the function pv, the author finds that, with decreasing temperatnrc and pressure, it does not increase indefinitely to a limiting value, bnt attains a msximum value at a temperature near 0" C., and then de- creases. If the vapour be removed from its point of saturation by expansion at a constant temperature, the product 2v.j is found first to increase, a t ordinary temperatures, to attain a maximum value at a certain point of dilution of the vapour, and then to decrease, whereas a t very low temperatures a progressive decrease from the point of saturation is observed.The deviation of diluted aqueous vnpour from Boyle's law is therefore essentially different from that of ordinary gases and vapours, and is rather of the nature observed by Mendelejeff in the case of rarefied atmospheric air. C. P. c. Abnormal Vapour-densities. By J. GUAR E s c H I ( A h . d.' Acad. cl. Eolog'za [3], viii, 193).-The author gives a general view of the experimental investigations which have any bearing on the .question of the so- called abnormal vapour-densities. He endeavoiirs to show that in most cases of abnormal vapour-densities a partial or total decom- position can be proved with more or less certainty, and therefore that they are not really exceptions to Avogadro's law.Some Properties of Boric Acid. By A. I) I T T E ( Co??zpt. retad., lxxxv, 1069--1072).-1n this paper the author has determined the heat disengaged by the hydration of boric anhydride, which he finds to be 6300 thermal units a t 14;' per one equivalent boracic acid to three of water. I n the case of solution of the hydrated acid an absorptioii of heat, amounting to 3181 thermal units per equivalent, takes place T. C.GENERAL AND PHYSICAL CHEMISTRY. 195 011 the formation of a saturated solution, the solution of tlic hydrous acid thus absorbing about half the heat disengaged by its hydra- tion.The author then gives tables of the specific gravity of the acid a t various temperatures, both in the hydrated and anhydrous condi- tion ; he also determines the coemcient of dilatation between 12" and 80' as 0.0014785, and between 12" and 60" as 0.0015429, amcl gives tables of the solubility of both kinds of acid a t different tcmperatures. I n conclusion he calls attention to the use t o which this action may be put as a lecture experiment to show evolution of heat by chemical action, 100 grams of the anhydride mixed with 125 grams of water being able to melt in a few minutes an ingot of Dtwcet's alloy. J. IT. T. Surface-tension of Aqueous Solutions of Alcohols and Fatty Acids. By 31. DUCLAUX (Compt. rend., lxxxv, lOG8--1069).-'l'he author states that, by allowing different solutions of alcohols and fattj- acids t o flow from a tube having an orifice of known diameter, under constant pressures, and counting the number of drops given by the different solutions, he can obtain the superficial tension of thcse solu- tions by a simple calculation.By comparing these tensions he arrives at the following result :-If solutions of difl'erent den& ties of alcohols or fatty acids having the same superficial tension are compared, the volume-percentages of alcohol or acid whicli they contain have a constant ratio independent of the tension. Tlius, let x be the per- c:entage of alcohol or acid in a liquid, the superficial tension of which = y, and let TC = f (y) be the equation representing the curve of the tensions for a given substance, then x = kf (y) will be the equation of the same curve for any other substance ; or, the function of 1~ in the above expression is the same f o r all bodies of the same organic series, and is modified only from one to the other by the introduction of a constant coefficient, k , which characterises each body.J. M. T. On the Capillary Angle and the spreading out of Liquids upon Solids. By G. QUINCKE (Am. Phys. Chem. 121, ii, 145-194). --The author has published previous investigations upon the surface- tensions and capillary angles of liquids by different methods, but the result of one method did not agree with those of another. He has therefore adopted a direct measure of the angle of capillarity, by ob- serving the angle between the reflections of the same ray from the two surfaces near the dividing edge.He was in this way able t o make very accurate measures. The angle between the same substances was found to vary from different causes. That of a, drop in contact with a glass plate was less the greater the height of fall of the drop on to the plate. This is explained by the fact that if a drop is once spread out it does not contract again properly, and so makes a smallcr angle a j t h the plate than it would otherwise. But the most important modifjing cause was the greater or lesser cleanliness of the plate. The best method of cleaning a surface was to heat it in sulphuric acid, wash it, and allow it to stand in distilled water, and then t o dry it in the colourless flame of a Bunsen's burner.Upon a surlace thus pre- pared fluids like water, alcohol, &c., seemed to spread out a t once and to have a capillary angle of zero. But a few seconds sufficed for the196 ABSTRACTS OF CHEMICAL PAPERS. condensation of air or moisture on the surface and a consequent altem- tion of the angle. The longer the surface was exposed to the air tlie less clean it became and the greater was the angle. The slightest trace of oil was sufficient to affect the surface, and when once present was difficult to remove. It seems probable that the angle for liquids, such as water, alcohol, &c., upon clean glass, crystal, or metal surfaces is zero, and that the liquids immediately spread out, but that when i t has a different value, a layer of some substance is present, upon the solid surface. This layer may be excessively thin, too thin even to show the interference colours.It may consist of foreign solid, liquid, or gaseous substances, or part of the liquid which is being invcstigatetl may itself spread out over the surface, and form a very thin layer mitlt a different snrface-tension from the rest of the liquid which may then rest upon it in a lenticular form. The presence of Ghese layers may be proved by the so-called creeping of salts, or by their conduction of electricity. If two liquids, miscible in all proportions, are in contact with one another and with a third solid body, they will have no definite common surface with a surface-tension, and therefore the one which has the greater surface-tension at the solid surface will be driven by the other away from the solid.By this may be explained some of the phenomena, of diffusion of salts through membranes, &c. Studies on Chemical Volumes. By W. OSTWALD (Chew. Centr., 1877, 25-32 and 42--43).-Several attempts have been made to answer the question how two acids divide themselves towards a base in aqueous solutions. Berthelot and St. Martin adopted a chcrnical method (Ann. Ch+m Phys. [4], xxvi, 433, 1872), which, however, is open t o many objections. The calorimetrical method of A. Muller (Pogg. ATZ'Y~., Suppl. vol., vi, 123, 1875), and that of J. Thomsen, de- pending on the evolution of heat (Pogg. Ann., cxxxviii, 65, 1869), give much more satisfactory results. The method now proposed by the author depends on the measure- ment of the specific gravities of the solutions.Since alterations of volume generally take place during chemical processes in aqueous solutions, it follows that if these are different in one case from what they are in another, the relative magnitudes of action of two bodies acting simultaneously may be measured by the alteration in volume. For instance, the sp. gr. of an equivalent of N a P in solution is 104051, and that of an equivalent of SO3* in solution is 102970 com- pared with water at 20" ; hence the sp. gr. of NaO.SO, in solntion should be 207021. It is found to be 205218 ; hence a difference of - 1103. With NaO and NO5 the difference is found to be - 1868. They will account for Moser's pictures (Hanchbilder) . J. H. P. Thus- NaO.. ..........= 104Q51 sp. gr. NO, ........... = 103083 Sum.. ...... 207134 Na0.N05 found . . 205266 - 1868 * 0 = 8 : S = l G .INORGANIC CHENISTRY. 197 The difference between the contractions in the two cases aniounts to -7’765. I n this way a series of numbers is obtained which may be compared with those obtained by Thomsen by means of the formula The agreement of the numbers obtained by the shown in the following table :- Found by author’s method. NaO SO3.2SO3 ............ - 129 N&OSO3.+SO3 ............ - 213 NaOSO3. so3 ............ - 320 NaOXO3.2SO3 ............ - 398 NaOS03.4S03 ............ - 452 two methods is C:hdatcd b j iornllllu. - 132 - 213 - 309 - 396 - 463 n n + U.8 Thornsen’s formula is - x const. In other experiments Guldberg’s formula (Guldberg et Waage, Etudes szw les Afinite‘s c h i m i p e s , Christiania, 18767) is employed, and the results are shown to agree.The analogy between the change of volume and evolution of heat is very striking, as in the following cases :- Evolution of heal,. Condensation. Na0.S03-Na0.N0, ...... - 2072 - 765 Na0.S03--Na0.HC1 ...... - 1949 - 740 Na0.S03.$S03 ............ - 631 - 213 Na0.S03.+N05 ............ - 1292 - 472 Na0.S03.2N0, ............ - 2026 - 748 Na0.S03.2HC1.. .......... - 1878 - 688 K. Hofmnnn proposed a method somewhat similar to t,hat of the author to solve a similar question (Pogg. Ann., cxxxiii, 5 i 5 ) . G. T. A.189G e n e r a l a n d P h y s i c a l Chemistry.On some Points connected with the Chemical Constituentsof the Solar System. By J.H. G L A D S T O N E (Phil. Nus. [5], iv,379-385) .-The formation of the solar system by thc condensationof a nebulous mass made up of many different chemical elements, willhave resulted in a distribution of those elements dependent on the twofollowina considepations :-(1.) During the process of cooling the least volatile constituents willcondense first and sink towards the centre of gravity, while thc restwill arrange themselves inore or less in the order of their volatility.(2.) As mas pointed out by Mr. C. J. Stoneg, in an atmosphere de-creasing in temperature from within outwards, the lightest moleculeswill be the farthest from the centre of gravity.The most volatile elements, and those of least vapour-deiisity, willtherefore be the outermost in a hot nebulous mass.This distributionis seen to occur in the sun, where hydrogen forms the bulk of the outeratmosphere, mixed with small quantities of sodium and magnesium,while the vapoui* of iron is found only in a lower stratum of the sun’satmosphere, and platinum has not been detected at all.According to the nebular theory, the planets wcwe formed by theseparation of some of the outer portion of the original mass; weshould therefore expect them to contain a preponderance of the morevolatile elements, and those of least vapour-density ; and this is in-deed the case. Of thecnori-metallic elements, those which are plentifulhave an average vapour-density of 19.8, while those wliich are com-paratively rare have an average vapour-density of 63 : grouping themetals into four classes-plentiful, common, rare, and very rare, wefind that the average vapour-densities of each class are, 37.8, 104.5,106.7, and 122.9 respectively.The meteoric stones which fall to the earth from interplanetary spacesshow this preponderance of the lighter elements still more strikingly.There are, however, in both cases many exceptions to thc rule, for in-Etance, glucinum and lithium are both rare, while lead is w r y common ;such exceptions may be cxplained by supposing these elements to havebeen combined with others forming compounds of g1-catcr or lessvolatility than themselves ; thus carbon is very difficult to volatilise,but its compounds with oxygen, hydrogen, &c., are gases at the ordi-nary temperature.The heads of comets emit light giving band spectra which areusually referred to carbon ; the volatilised carbon of the electric lampwas also found to give these band spectra ; in cornets, however, thecarbon is probably combined with oxygen or hydrogen.The Influence of Temperature upon the Coefficients ofRefraction of the Natural Sulphates of Barium, Strontium,and Lead. By A.ARZRUNI (Jahrb. f. J&., 1877, 52t;--527).--Theauthor considered it desirable to ascertain whether the coefficients ofF. D. €3.VOL. XXXlV. 2190 ABSTRACT8 OF CHEMICAL PAPERS.refraction remained constant or varied with the temperature a t whichthe determination was made. Three isomorphous sulphates, viz.,barium salpliate, strontium sulphate, and lead sulphate were selectedfor the investiption ; firstly, because Descloiseaux observed in them agreat variation of the angle of the optical axis with the temperature,from which it might justly be inferred that any change in the co-efficient of refraction could easily be determined ; and, secondly, itwas desirable to ascertain if isomorphous cornpounds with analogousoptical characteristics still remained analogous to an increase oftemperature.F o r these observations the author prepared prisnis outof a barytes crystal from Dufton, a cmlestine crystal from Lake Erie,and an anglesite crystal from Monte Poni, the results of the investiga-tions being briefly as follows, vie. :-(1.) The principal coefficients ofrefraction of the above-mentioned isomorphous sulphntes differ fromeach other with the temperature, but all of them decrease with anincrease of temperature.(2.) The decrease in the coefficient of re-fraction of the three sulphates is an analogous one, and can he ex-pressed thus, viz., Further, y approaches the two others,whilst u withdraws from 6. (3.) With anglesite the refraction isinversely as the temperaftme, whilst the dispersion increases for dif-ferent colours. (4.) The directions of the maximum, medium, andminimum expansicjn of the three compounds by heat do not staid inany relationship to the values of the directions of optical elasticity inthem, or to the alteration of the velocity of light in these three direc-tions. C. A. B.> u > 6.On the Law of Absorption and its Employment in Quantita-tive Spectrum Analysis.By G. Govh (Conyt. r m r l . , lxxxv, 1046-1049).-The author at first alludes to the phenomena of absorption,and compares tlie increase and dilatation o€ absorption-bands ‘‘ due tothickening in the absorbing medium,” to the increase in the brightlines in the spectra of incandescent gases caused by increase in thetemperatiwe aEd pressure. He says tliat it is not possible to determinethe absorbing power of a body, unless the coefficients of absorptionare known f o r all wave-lengths which can be studied. He thinks,however, tlhatj the absorbing power of bodies may be determined, eitherby direct comparison of the curves of chroniatic absorption, or bymeasuring tlie intensity of the light along the whole length of thespectrum. To use the first of these methods, lie employs the ab-sorbent in tlie shape of a prism, and by placing one of its plane facesagainst the slit, with the centre angle touching one end of it, he ob-tains a gradually increasing thickness of the absorbing mediumthroughout the whole leiigth of the slit, any deviation heing ncu-tidised by a prism of the smallest absorbing power.When whitelight is passed through the arrangement described above, the spectrumshows more or less wavy shadows, representing to the eye the lawaccording to which tlie coefficient of absorption of the medium varieswith the wave-length of the incident light. Should the slit be dividedlongitudinally into two equal parts, i t is easy t o compare the twochromatic absorption-spectra produced.By using solar light theabsorption-curves may be compared with E’raunhof er’s lines. ThGENERAL AND PHYSICAL CHEMISTRY. 1'31author concludes by saying that in cases requiring great accuracy &ismethod is not' applicable. J. M. T.Theory of the Action of certain Organic Substances inincreasing the Sensitiveness of Silver Haloids. By M. C. L E A(Amer. J. of Xci. [3], xiv, 96-99).-Explanations of this action ofcertain organic substances have been offered by Poitevin and by H.Vogel, who suggest that' it' is due to their af€inity for the halogen.The author, however, .points out that all the organic suhstances pos-sessing the property in question are reducing agents, and argues that.i t is their affinity for oxygen which, aiding as it does the affinity of tliehalogen for hydrogen, determines the decomposition of the silver-salt.That this is so is provcd by the fact that when pyrognllol is acided torecently precipitated silver iodide, and the mixture exposed t o sun-light, it exhibits a distinctly acid reaction at tlie end of 15 minutes.Had the first-mentioned explanation been the true one, an iodo-substi-tution-product would have been formed, but no acid.Substances which have a great affinity for iodine, as potassinmcarbonate and starch, do not increase the sensitiveness of the silverhaloid.These facts support tlie views of the author, as does also theprocess of alkaline development, tlie alkali, 11y ncutralising the acidproduced, assisting the development of the image.Y. D, B.A Battery in which the Carbon Electrode is the one At-tacked. By P. JABLOCHKOFP (Conzpt. rend., lxxxv, 1052-1053).-The electricity produced by electro-magnetic machines is due t o thecombustion of carbon. The author has attempted to produce electricityby direct action on carbon. As carbon is not attacked by liquids atjordinary temperatures, he has constructed a hot liquid electro-chemicalbattery. For this purpose lie uses either potassium or sodium nitratein which ordinary coke is used as one electrode, and for thc otherplatinum-iron, or any metal not attacked by the liquid in lwesenceof carbon. By the addition of various metallic salts the electro-motive force may he modified, the nietals being deposited on theunattacked electrode.The electromotive force of the battery theauthor states to vary between 2 and 3 uiiits, while the Runscn batterygives a maximum of 1.8, and Gernet 2.1. To start the action of thebattery, a piece of iiicandescent coke is placed in contact with thecrushed nitrate, heat) is pyoduced, and the action commences. A largequantity of carbon dioxide and other gases are evolved, which theauthor proposes to use as a motive power. A description of the appa-ratus is given. J. 31. T.The Movements of Electrified Mercury. By HE R M AS NHE R W I G (8w.n. Ph?ys. Chem. [Z], i, 73--95).-Earlier experiments onthe capillary depression of electrified mercury (this Journal, 1877, i,677) had rendered it probable that the surface of the glass tube inwhich the mercury was contained was attacked. I n the experimentshere detailed the mercury contained in a capillary tube vas electrifiedby connecting it with one pole of a Holtz machine, while the otherpf192 ABSTRACTS OF CHEMICAL PAPERS.pole was connected with the earth, sparks being meanwhile allowed topass between the poles which were not too far apart.When the mercury was connected with the positive pole, i t rorcgradually in the tube when the machine was set in motion, forming, a tthe highest point to which it reached, and where the surface remainedfor some time, a dirty ring, which was not removed by repeatedlypouring dean quicksilver through the tube ; this ring held up thethread of mercury when it was no longer electrified.When, on theother hand, the mercury was connected with the negative pole of themachine, it did not rise so high in the tube, no ring was formed unlessthe electrification was continued for a long time, and on its ceasing thequicksilver sank immediately to its former level ; this latter fact clearlyshows that the cohesion of the quicksilver is diminished by electrifica-tion.Drops of mercury placed upon a horizontal glms plate, and posi-tively electrified, rapidly formed a ring of dirty mercury on the plate :when, however, both the plate and the metal hid been carefully dried,these rings were only obtained with great difficulty. The quicksilveris therefore oxidised if water he present. That the diminution of therapillary depression is not caused by such oxidation alonc is, liowever,shown by the fact that when tm-o similar tubes were filled, the onewith hot, dry quicksilver, the other with the same substance inten-tionally moistened, both tubes being closed at the top, the S~TIIC phe-nomena were observed in each on electrification. Furtlicr, when thespace above the quicksilver -as filled with dry hydrogen, the sameeffects were produced as with air, though not so rapidly, although allchance of oxidation was removed.When the mercury of a barometer is positively electrified in thcsame manner, the discharge from thc upper surface being facilitatedby connecting the outer surface of the glass with the ground, it ap-pears to boil violently, and glowing particles are projected against thesides of the tube ; the effect is much inferior with negative elcctricity.The diflerent action of the two electricities is explained by sup-posing that where a break occurs in a series of conductors, the nega-tive electricity flows more easily, while the positive collects on thesurface till a far greater tension has been attained, thus occasioning,on at length passing off, a much greater disturbance.When the mercury is connected with the positive pole the entire1-acuous space is filled with bluish-green light, which, examined withthe spectroscope, is shown t o be mercury light ; close to the mercnrymeiiiscus a yellowish-green fluorescent light is also observed whichgives an almost continuous spectrum ; when connected with the nega-tive pole the bluish-green light is less developed, while the fluorescentlight fills the upper part of the vacuum.The above liypothesis cx-plains these facts also, since the negative electricity passing moreeasily from the quicksilver surface, the charge would accumulate a tthe upper portion of the tube where its presence would be indicatedhy the fluorescent light, whereas the positive would be found close tothe surface of the mercury.The same explanation applies to the movement of threads of mer-cury in horizontal capillary tubes : it was found, for instance, thaGENERAL AND PHYSICAL CHEMISTRY. 193when a platinum-wire, connected with the negative pole of the machine,was placed near one end of the thread, the merciiry rapidly approachedthe wire, whereas i&en it was connected with the positive pole themercury was almost always repelled.0 bservations made with tubescontaining threads of mercury bounded a t each end by water, wereattended with like results.The results o€ the experiments already made are summed up asfollows :-(1.) Powerful charges of electricity diminish the cohesion of mer-cury more than the adhesion between mercury and glass, thus lessen-ing the capillary depression in glass tubes.(2.) When positive electricity a t high tension escapes from the sur-face of mercury in glass vessels, tlie surface being exposed to the air,oxidation takes place, more especially when moisture is present. Nega-tive electricity has a reducing action.(3.) The passage of electricity (particularly positive) at high tensionfrom mercury to glass occasions decomposition of the glass.(4.) The three facts above mentioned are the united cause of theobserved diminution of the capillary depression.( 5 .) Negative electricity escapes from highly charged conductors ata less tension than positive.(6.) When surfaces of mercury are charged with electricity, theformation of mercury-vapour is greatly facilitated. Positive electricityhas by far the greatest effect.(7.) Mercury-vapour is relatively an excellent conductor of elec-trici ty . F. I). B.On the Specific Heat of Vapours and its Variations withthe Temperature. By E. WIEDEMANN (Am. Phys. C h n . [2], ii,1'35-21 7) .-In Regnault's researches upon specific heats of vapours,he determined the specific heats at high temperatures by finding theqiiantities of heat given up in each case in condensing the gas fromtwo different high temperatures to the same lower temperature, thedifference between the two quantities giving the specific heat betweenthe two high temperatures.He found that the specific heat et a t atemperature t, might be represented by the formula c, = c, + gat,when c, is the specific heat a t 0" and 2u the alteration of specific heatfor one degree of temperature. His method is objectionable in this :that though the amount of heat given up in each case may be largeand capable of determination with only a small proportiorlate error,nevertheless a small error may bear a, large proportion to the dif-ference between two determinations, and so considerably affect theresult.Wiedemann avoids this difficulty by using an arrangementin which he can obtain the vapours a t low pressures and therefore atlower temperatures, and he observes directly the heat given off incooling a vapour from a higher to a lower temperature. He hasinvestigated thus the values of c, and u in Regnault's formula forchloroform, ethyl bromide, benzene, acetone, acetic ether, and ethyloxide. His resulting specific heats differ from those of Begnault byfrom 3 to 5 per cent., but agree much more closely among them194 ABSTRACTS OF CHEMICAL PAPERS.selves, the variations from Regnault's results being probably clue toi 111 purities .He finds in general that the greater the specific heat of a 1iqu:d thflgreater is that of its vapoin.The coefficient a is o€ tlic same order ofmagnitude for the liquid and its vapour, and in some cases nrarly thesame for the two, but it varies greatly for dif-ferent vapours.J. H. P.The Internal Condition and Latent Heat of Vapours.. ByP. C. PUSCHL (Chem. Centr., 1877, p. 318).-Bg a metliod incle-pendent of the second law of thermodynamics, the author deduces thegeneral equation which holds for saturated vapours. Hc furthershows that, in the cycle of operations upon a mixture of liquid andrapour, which consists (I), in allowing the same to expand at a con-stant temperature; ( 2 ) ' heating a t the rolumc attained and keptconstant; (3), allowing it to contract to its initial volume a t thisincreased temperature; and (4), allowing it to cool to the init'ialtemperature while maintaining its volinme constant at t,his point,-thework expended is not the equivalent, but is in excess of the heat ob-tained.There has therefore been a gain in internaJ work, the qnnntitgof which may be determined in the case of water and its vapour fromRegnault's experimental data. The values of the forces which deter-mine the volume of water vapour, under the external pressure, mayeasily be determined for temperatures between 0" and 200". Respect-ing the function pv, the author finds that, with decreasing temperatnrcand pressure, it does not increase indefinitely to a limiting value, bntattains a msximum value at a temperature near 0" C., and then de-creases.If the vapour be removed from its point of saturation byexpansion at a constant temperature, the product 2v.j is found first toincrease, a t ordinary temperatures, to attain a maximum value at acertain point of dilution of the vapour, and then to decrease, whereasa t very low temperatures a progressive decrease from the point ofsaturation is observed. The deviation of diluted aqueous vnpour fromBoyle's law is therefore essentially different from that of ordinarygases and vapours, and is rather of the nature observed by Mendelejeffin the case of rarefied atmospheric air. C. P. c.Abnormal Vapour-densities. By J. GUAR E s c H I ( A h . d.' Acad.cl. Eolog'za [3], viii, 193).-The author gives a general view of theexperimental investigations which have any bearing on the .questionof the so- called abnormal vapour-densities.He endeavoiirs to showthat in most cases of abnormal vapour-densities a partial or total decom-position can be proved with more or less certainty, and therefore thatthey are not really exceptions to Avogadro's law.Some Properties of Boric Acid. By A. I) I T T E ( Co??zpt. retad.,lxxxv, 1069--1072).-1n this paper the author has determined theheat disengaged by the hydration of boric anhydride, which he findsto be 6300 thermal units a t 14;' per one equivalent boracic acid to threeof water. I n the case of solution of the hydrated acid an absorptioiiof heat, amounting to 3181 thermal units per equivalent, takes placeT. CGENERAL AND PHYSICAL CHEMISTRY.195011 the formation of a saturated solution, the solution of tlic hydrousacid thus absorbing about half the heat disengaged by its hydra-tion. The author then gives tables of the specific gravity of the acida t various temperatures, both in the hydrated and anhydrous condi-tion ; he also determines the coemcient of dilatation between 12" and80' as 0.0014785, and between 12" and 60" as 0.0015429, amcl givestables of the solubility of both kinds of acid a t different tcmperatures.I n conclusion he calls attention to the use t o which this action may beput as a lecture experiment to show evolution of heat by chemicalaction, 100 grams of the anhydride mixed with 125 grams of waterbeing able to melt in a few minutes an ingot of Dtwcet's alloy.J.IT. T.Surface-tension of Aqueous Solutions of Alcohols and FattyAcids. By 31. DUCLAUX (Compt. rend., lxxxv, lOG8--1069).-'l'heauthor states that, by allowing different solutions of alcohols and fattj-acids t o flow from a tube having an orifice of known diameter, underconstant pressures, and counting the number of drops given by thedifferent solutions, he can obtain the superficial tension of thcse solu-tions by a simple calculation. By comparing these tensions he arrivesat the following result :-If solutions of difl'erent den& ties of alcoholsor fatty acids having the same superficial tension are compared, thevolume-percentages of alcohol or acid whicli they contain have aconstant ratio independent of the tension.Tlius, let x be the per-c:entage of alcohol or acid in a liquid, the superficial tension of which= y, and let TC = f (y) be the equation representing the curve of thetensions for a given substance, then x = kf (y) will be the equationof the same curve for any other substance ; or, the function of 1~ in theabove expression is the same f o r all bodies of the same organic series,and is modified only from one to the other by the introduction of aconstant coefficient, k , which characterises each body. J. M. T.On the Capillary Angle and the spreading out of Liquidsupon Solids. By G. QUINCKE (Am. Phys. Chem. 121, ii, 145-194).--The author has published previous investigations upon the surface-tensions and capillary angles of liquids by different methods, but theresult of one method did not agree with those of another.He hastherefore adopted a direct measure of the angle of capillarity, by ob-serving the angle between the reflections of the same ray from the twosurfaces near the dividing edge. He was in this way able t o makevery accurate measures. The angle between the same substances wasfound to vary from different causes. That of a, drop in contact with aglass plate was less the greater the height of fall of the drop on to theplate. This is explained by the fact that if a drop is once spread outit does not contract again properly, and so makes a smallcr anglea j t h the plate than it would otherwise. But the most importantmodifjing cause was the greater or lesser cleanliness of the plate.The best method of cleaning a surface was to heat it in sulphuric acid,wash it, and allow it to stand in distilled water, and then t o dry it inthe colourless flame of a Bunsen's burner.Upon a surlace thus pre-pared fluids like water, alcohol, &c., seemed to spread out a t once andto have a capillary angle of zero. But a few seconds sufficed for th196 ABSTRACTS OF CHEMICAL PAPERS.condensation of air or moisture on the surface and a consequent altem-tion of the angle. The longer the surface was exposed to the air tlieless clean it became and the greater was the angle. The slightesttrace of oil was sufficient to affect the surface, and when once presentwas difficult to remove. It seems probable that the angle for liquids,such as water, alcohol, &c., upon clean glass, crystal, or metal surfacesis zero, and that the liquids immediately spread out, but that when i thas a different value, a layer of some substance is present, upon thesolid surface.This layer may be excessively thin, too thin even to showthe interference colours. It may consist of foreign solid, liquid, orgaseous substances, or part of the liquid which is being invcstigatetlmay itself spread out over the surface, and form a very thin layer mitlta different snrface-tension from the rest of the liquid which may thenrest upon it in a lenticular form. The presence of Ghese layers may beproved by the so-called creeping of salts, or by their conduction ofelectricity.If two liquids, miscible in all proportions, are in contact with oneanother and with a third solid body, they will have no definite commonsurface with a surface-tension, and therefore the one which has thegreater surface-tension at the solid surface will be driven by the otheraway from the solid.By this may be explained some of the phenomena,of diffusion of salts through membranes, &c.Studies on Chemical Volumes. By W. OSTWALD (Chew.Centr., 1877, 25-32 and 42--43).-Several attempts have been madeto answer the question how two acids divide themselves towards a basein aqueous solutions. Berthelot and St. Martin adopted a chcrnicalmethod (Ann. Ch+m Phys. [4], xxvi, 433, 1872), which, however, isopen t o many objections. The calorimetrical method of A. Muller(Pogg. ATZ'Y~., Suppl. vol., vi, 123, 1875), and that of J. Thomsen, de-pending on the evolution of heat (Pogg. Ann., cxxxviii, 65, 1869),give much more satisfactory results.The method now proposed by the author depends on the measure-ment of the specific gravities of the solutions. Since alterations ofvolume generally take place during chemical processes in aqueoussolutions, it follows that if these are different in one case from whatthey are in another, the relative magnitudes of action of two bodiesacting simultaneously may be measured by the alteration in volume.For instance, the sp. gr. of an equivalent of N a P in solution is104051, and that of an equivalent of SO3* in solution is 102970 com-pared with water at 20" ; hence the sp. gr. of NaO.SO, in solntionshould be 207021. It is found to be 205218 ; hence a difference of- 1103. With NaO and NO5 the difference is found to be - 1868.They will account for Moser's pictures (Hanchbilder) .J. H. P.Thus-NaO.. .......... = 104Q51 sp. gr.NO, ........... = 103083Sum.. ...... 207134Na0.N05 found . . 205266- 1868* 0 = 8 : S = l G INORGANIC CHENISTRY. 197The difference between the contractions in the two cases aniounts to-7’765. I n this way a series of numbers is obtained which may becompared with those obtained by Thomsen by means of the formulaThe agreement of the numbers obtained by theshown in the following table :-Found byauthor’s method.NaO SO3.2SO3 ............ - 129N&OSO3.+SO3 ............ - 213NaOSO3. so3 ............ - 320NaOXO3.2SO3 ............ - 398NaOS03.4S03 ............ - 452two methods isC:hdatcd b jiornllllu.- 132- 213- 309- 396- 463nn + U.8Thornsen’s formula is - x const.In other experiments Guldberg’s formula (Guldberg et Waage,Etudes szw les Afinite‘s c h i m i p e s , Christiania, 18767) is employed, andthe results are shown to agree.The analogy between the change of volume and evolution of heat isvery striking, as in the following cases :-Evolution of heal,. Condensation.Na0.S03-Na0.N0, ...... - 2072 - 765Na0.S03--Na0.HC1 ...... - 1949 - 740Na0.S03.$S03 ............ - 631 - 213Na0.S03.+N05 ............ - 1292 - 472Na0.S03.2N0, ............ - 2026 - 748Na0.S03.2HC1.. .......... - 1878 - 688K. Hofmnnn proposed a method somewhat similar to t,hat of theauthor to solve a similar question (Pogg. Ann., cxxxiii, 5 i 5 ) .G. T. A