SCIENTIFIC AND ANALYTICAL

CHEMISTRY.

On the Analysis of Mineral Phosphates, by R. WARINGTON, jun., F.Č.S., Assistant to Professor CHURCH, Royal Agricultural College, Cirencester.

THE minerals which are at the present time chiefly resorted to as sources of phosphoric acid are apatite, guano, and coprolite. The analysis of some of these, as of the purer forms of apatite, and of Peruvian guano, presents but little difficulty, at least when the commercial value of the mineral is the object in view, but when we come to substances like coprolite, or the curious altered guano known as sombrerite, the case is different. These minerals generally contain a considerable amount of oxide of iron and alumina, besides lime, magnesia, &c., and we have to solve the serious problem of the estimation of phosphoric acid in the presence of these bases, as well as their separation from each other.

On consulting standard manuals of analysis, it is presently seen that very few of the processes there recommended for the determination of phosphoric acid are here applicable. The molybdic acid method is inad missible from the large amount of phosphoric acid to be determined. The mercurial method fails us owing to the presence of alumina. The uranium method is equally unavailable from the presence of iron, or can be employed only if the iron be previously reduced by means of protochloride of uranium. The tin method is free from all these objections, and is no doubt, when carefully conducted, an excellent process, but the time it takes up is considerable, and it appears for other reasons to be unfitted for very general use.

Different methods have been employed by individual chemists, but none have proved so eminently satisfactory as to be extensively adopted. In this state of the case, the result of some experiments recently made in this laboratory may not be unacceptable to the agricultural analyst.

a little acetate of ammonia, otherwise the filtrate is apt to be turbid.

There are several ways of treating the precipitate: it may be dissolved in nitric acid (which, for this purpose, must not be too weak), the solution diluted, and the lead precipitated by means of sulphuretted hydrogen; or the nitric solution may be treated with an excess of sulphuric acid, and the lead precipitated as sulphate. The first plan is unexceptionable, and yields excellent results; the necessary dilution entails, however, a subsequent concentration, which occupies some time. If acetate of lead has been used, perhaps the best plan for ordinary purposes is to treat the phosphate of lead with oxalic acid and a few drops of oxalate of potash. The decomposition is rapid and complete; the oxalate of lead may, after a short time, be separated by filtration. Oxalic acid does not perfectly decompose the highly aggregated precipitate obtained with nitrate of lead and litharge. The action of sulphuric acid on the precipitate is incomplete, whether nitrate or acetate of lead has been employed. Oxalate of lead is nearly insoluble in oxalic acid; its solution, if treated with sulphuretted hydrogen water, appears only very slightly discoloured when looking through a depth of several inches.

The lead being separated, the solution now contains the whole of the phosphoric acid, with a little iron; some citric acid is added, and an excess of ammonia; the clear solution is finally treated with magnesia mixture, and the phosphoric acid separated in the usual way.

The lime, magnesia, and alkalies are readily determined in the original filtrate from the phosphate of lead, the excess of lead being first precipitated by means of sulphuretted hydrogen. A little citric acid must be

added before the solution is made ammoniacal for the determination of magnesia, to hold in solution the oxide of iron and alumina.

of the convenient determination of all the bases except This method has the advantage of speed, and admits iron and alumina; its accuracy has been carefully tried by operating on artificial mixtures of known composition. The results were very satisfactory.

The next method we have to describe is an old one.

The nitric acid solution of the mineral, previously freed from silica in the usual way, is treated cautiously with dilute ammonia to remove all unnecessary excess of acid, the clear liquid is then treated in one of two ways; either an excess of neutral acetate of lead is added, or the solution is treated with nitrate of lead, and digested with successive portions of finely-powdered litharge till slightly alkaline, the liquid in this case being finally acidified with a few drops of acetic acid. The former plan is generally to be preferred, though the precipitate is considerably more bulky, as only a portion of the iron, and probably none of the alumina, is in this case precipitated with the phosphoric acid.

The precipitated phosphate of lead is warmed for some minutes to induce aggregation, and then thoroughly washed by decantation, the washings being filtered. The washing water should be slightly warm, and contain

When nitric acid alone is used to effect the solution of a coprolite a trace of phosphoric acid is sometimes left with the silica; it is therefore safest to dissolve in hydrochloric acid, and after the evaporation

to dryness to redissolve with nitric acid.

VOL. X. No. 239.-JULY 2, 1864.

is cautiously treated with dilute ammonia till a slight The acid solution of the mineral after separating silica, permanent opalescence is produced; a drop or two of oxalic acid is then added, and the fluid allowed to stand; the opalescence will disappear, and the solution, if iron ammonia is now added, and the fluid warmed for some is present, will become yellow. An excess of oxalate of time to aggregate the oxalate of lime, which is then citric acid added, and ammonia in excess, the phosphoric acid is then precipitated with magnesia mixture.

collected on a filter. The filtrate is concentrated, some

The oxalate of lime obtained will be found quite free from phosphoric acid; its amount, however, will be a little below the truth, from the partial solubility of oxalate of lime in oxalic acid. If to avoid this error, we attempt to neutralise the liquid before collecting the lime, phosphate of iron and alumina fall with the precipitate. The deficiency of lime amounted in the experiof phosphate of magnesia is not, however, in excess to ments to rather less than half a per cent.; the precipitate

that amount, as the citric acid not only holds in solution oxide of iron and alumina, but also, to some extent, oxalate of limet. A trace of lime will, nevertheless, always be found in the magnesia precipitate, not, however, sufficient to vitiate the result for any ordinary purpose, if the operation has been properly conducted.

Citric acid should in every case be preferred to tartaric, on account of the superior solubility of its compounds with magnesia‡.

CHEMICAL NEWS, July 2, 1864.

phosphoric acid. § The oxide of iron is to be thoroughly washed by decantation, redissolved, precipitated by ammonia, collected, and weighed. The results are accurate. Lime and magnesia may, if it is wished, be determined in the filtrate from the phosphate of iron.

The processes we have now mentioned have all been carefully tested, both qualitatively and quantitatively. Many of the reactions treated of are probably quite familiar to the readers of the CHEMICAL NEWS; the detailed description has, however, seemed necessary to give a complete view of the subject. In conclusion, I must express my obligations to Mr. C. Jacobsen, a graduate of the College, for his valuable assistance in these experiments.

avoided.

We turn now to the second part of our subject, the determination of alumina and oxide of iron; this has to be effected in a separate analysis. One of the advantages promised by the tin method already referred to, is, that it admits of the determination of all the bases in one portion. We find, however, a hint in Fresenius that if much iron is present, a portion of this is held by the binoxide of tin. In experiments made with this method 3 per cent. only of oxide of iron was obtained from a coprolite known to contain 5 per cent. ; the alumina was apparently unaffected. This is particularly unfortunate, as the separation of phosphoric acid by the tin leaves the bases all as nitrates, and consequently admitting of easy

separation.

Alumina is best determined by precipitating the acid solution with excess of caustic soda, digesting for some time, and finally separating the clear fluid by decantation and filtration. The solution contains the alumina plus phosphoric acid. The phosphoric acid is easily removed by treating cautiously with chloride of barium till it ceases to produce a precipitate. A little carbonate of soda is then added to remove the excess of baryta; lastly, some more caustic soda. The whole is warmed and filtered. The alumina is separated from the filtrate in the usual way.

The original precipitate by soda contains the whole of the iron. This is best determined by the volumetrie method with permanganate of potash. If, however, a gravimetric determination is desired, we proceed as follows:-The precipitate containing the mixed oxide of iron and phosphate of lime is dissolved in hydrochloric acid, a slight excess of ammonia added, and finally a considerable excess of acetic acid. The duid is gently warmed for a few minutes. The precipitated phosphate of iron is then thoroughly washed by decantation. Thus obtained, the phosphate of iron always contains a portion of phosphate of lime. Nearly the whole of this may be got rid of by redissolving the precipitate and treating with ammonia and acetic acid as before. The precipitate may then be collected; its formula is Fe2O, PO,. The following is perhaps a more strictly accurate method of procedure. The washed phosphate of iron is dissolved in a small quantity of oxalic acid, a little oxalate of potash added, and the whole warmed. Any oxalate of lime formed is separated by filtration, and the filtrate boiled for some time with a considerable excess of caustic potash. The precipitate consists of pure oxide of iron, the potash having removed the whole of the

+ Spiller. Quart. Journ. Chem. Soc., x. 110.
Jour. Chem. Soc., I. 304.

5

Researches on Oxygen, by Dr. G. MEISSNER.

title; the contents are so important as to deserve more than an ordinary review. We propose, therefore, to give as concisely as possible an abstract of parts of the work, so that our readers may have clearly before them the views of the author, and the experiments by which they are supported.

We may pass over the introduction-in which the author gives a rapid sketch of the history of Schonbein's discoveries, and the researches of other chemists on ozone —and pass at once to the chapter on "Electrised Oxygen."

The apparatus used by Dr. Meissner was a modification of an instrument by Siemens, recommended by Von Babo. It is constructed as follows:-Twelve very fine copper wires, about five decimetres long, are each inserted into a very thin glass tube a little longer than the wire, and about 0.33 millimetres in diameter. Each of these tubes is sealed at one end, and at the other end is fused a thin platinum wire, which is twisted with the copper wire within the tube, and projects some centimetres outside the tube. The twelve tubes thus made are arranged within a larger glass tube seven millimetres wide and six decimetres long, so that the projecting platinum wires of six of them are at one end, and those of the other six are at the other end of the wide tube. These two sets of wires are each twisted about a larger platinum wire, which passes through and is fused into the wall of the wide tube. The tubes of the one bundle are distributed as equally as possible among those of the other, and placed in close contact, so that the spaces surrounding them may be as narrow as possible. [An excellent drawing of this apparatus accompanies the book.]

When the ends of these larger platinum wires are connected with the secondary coil of a powerful induction apparatus, the discharge takes through the walls of the smaller tubes, and through the air which surrounds them. The passage of the electricity takes place without sparks, and with very little noise. On approaching. the ear only a faint crackling noise is heard. In the dark the bundle of small tubes is seen to be illuminated through its whole length with a reddish-violet light. It is obvious that in the above-described apparatus platinum wires may be used through the whole system.

By the passage of the electricity the air surrounding the small tubes is strongly ozonised, and by means of a suitable arrangement this ozonised air may be removed for examination and fresh supplied. The author effected this by the pressure of a gasometer.

Northcote and Church. Quart. Jour. Chem. Soc., vi., 53. "Untersuchugen über den Sanerstoff;" von Dr. G. Meissner, Professor in Göttingen. Hanover. Hahn'sche, Hofbuchandlung. 1863.

In the course of the experiments it was found to be necessary that the air should be perfectly dried before it was electrised. Some difficulty was experienced in accomplishing this, but the author succeeded at last by using next the gasometer a wide tube over a metre long, filled with pieces of chloride of calcium, and then two or three tubes together at least one or one and a-half metres long, filled with coarse glass powder drenched with English sulphuric acid. The perfectly dried air, after passing through the ozoniser, was submitted to reagents in receivers of glass connected with the ozoniser by means of mercury joints, this metal being unaffected by dry ozone. [This part of the apparatus is also figured in the book.]

The first point the author set about investigating was whether, besides the formation of ozone, any other changes were produced in oxygen by the passage of electricity. He found that when the ozonised air was passed through a strong solution of iodide of potassium it was deprived of every trace of ozone; but when this deozonised air was passed through pure water, it reappeared as a thick white mist, sometimes so thick as to render the surface of the water quite opaque. The mist resembles that formed by the cooling of steam. Temperature seemed to have no perceptible effect on the mist, as it was formed equally when the air was at 35° and o° C. It appeared also when the air passed merely through a moistened tube, and sometimes when it escaped from a weak solution of iodide of potassium. But when the solution of iodide was concentrated, and especially when the air was afterwards passed through a chloride of calcium tube, no mist appeared until the air again came in contact with water, showing that a certain amount of aqueous vapour is necessary to its

formation.

The appearance of the mist ceases directly the induction apparatus ceases to work, and is denser or lighter according as the electrical action is vigorous or weak. It is seen just as well when pyrogallic acid or other deozonising agents are used in place of iodide of potassium, and moreover is formed when the dry electrised air comes at once in contact with water.

Pure oxygen, however procured, when electrised in the same way, gave the same appearance, while hydrogen and nitrogen suffered no change. The author was thus led to the conclusion that when oxygen is electrised another modification is produced simultaneously with ozone, which he naturally concluded was Schönbein's antozone. He failed at first to establish the identity, and therefore gave the name atmizone (arμíšw—I smoke or fume) to this smoke-forming modification of oxygen. In the end, however, his researches left no doubt of the identity of atmizone with antozone.

(To be continued.)

A SIMILAR question has often been asked respecting ammonia, and the answer to the one will have an obvious bearing on the answer to the other.

It will be admitted that the finding of different compounds in which an equivalent of the hydrogen is replaced by the same monatomic element will settle the point in the negative.

Such cases are easily found. Substitution by bromine furnishes one example.

the action of bromine and phosphorus upon methylic alcohol. Any doubt that we might have about the real boiling point of the bromide of methyl will be removed by the consideration that the boiling point of bromide of ethyl is well known (40° C.), and that 13° C. is just about the calculated boiling point for the corresponding methyl compound.

A compound isomeric with bromide of methyl was obtained by Bunsen from the brom-hydrate of kakodylic acid. This compound is a gas. It condenses at about - 17° C.

Chlorine substitution also gives different compounds. Thus, chloride of methyl obtained by means of methylic alcohol, chloride of sodium, and sulphuric acid is different from the body obtained by acting upon marsh gas by means of chlorines.

Bayer has found the chloride of methyl prepared according to the former process dissolves in water at 14° C. to the extent of 4172 times (i.e., one volume of water dissolves 4°172 volumes of the gas).

The compound prepared from marsh gas, on the other hand, is dissolved only to the extent of 0.08 at 14° C. Moreover, the former of these bodies forms a crystalline hydrate with water, the latter none. (See Bayer's paper.)

We are, therefore, forced to admit that the hydrogen in the most simple hydrocarbon has not all of it the same function; and since there is an isomer of the bromide and of the chloride of methyl, we should expect to find an isomer of methyl alcohol itself.

London Institution, June, 1864.

On the Determination of Solubilities,*
by FRANK H. STORER.

THE term "solubility" is to be taken in its most comprehensive sense. I have no intention of attempting a strict definition of the word, or of discussing the forces upon which solution may depend. In the present state of science, the collection of experimental data, and the study and comparison of well-authenticated special observations, seem to be of far greater importance than the disputes of the earlier chemists whether the phenomena in question should be referred to the domain of chemical affinity, or be studied as a purely physical problem.† It need only be remarked that I am accustomed to class among phenomena of solubility all those reactions of liquids upon solid bodies, and those combinations of liquids with liquids excluding, for the present, molten metals and other substances in a state of igneous fusion -in which the chemical force as understood by Berzelius, for instance, is not the principal, and, as it were, overwhelming force in action; we may have, perhaps, "solution" depending upon merely physical forces, like adhesion or cohesion, and also upon these forces plus a certain amount of chemical force. It can, indeed, hardly

* Extracted from the Preface to the "First Outlines of a Dictionary of the Solubilities of Chemical Substances."

+ Dans les sciences naturelles, et surtout dans la chimie, les géné ralités doivent résulter de la connaissance minutieuse de chaque fait, et non la précéder.-Gay-Lussac, Premier Memoire sur la Dissolu

4

we

admit of a doubt that the chemical force is exerted in many cases of solution, while, at the same time, other forces unquestionably come into play, in which connection the old adage that "like dissolves like " should be borne in mind. Hence, while the manifestations of chemical affinity proper, as evinced by the combination of bodies in simple and definite proportions, constitute the main subject of chemical text-books, many of the less clearly defined phenomena of chemical science may fairly come within the scope of a treatise on solubilities. Thus, though in the term "solubility of a substance ordinarily include only the comportment of the substance towards water, alcohol, wood spirit, ether, oil of turpentine, benzin, and analogous hydrocarbons, and the other "neutral solvents," it is obviously sometimes proper to add observations on the action of acids and alkalies; for example, any account of the solubility of nitrate of baryta would be manifestly incomplete without a statement of the fact that this salt is taken up but sparingly by nitric acid. Again, in the solution of chloride of silver in ammonia water, and that of various salts-as sulphate of lime, for example, in acids-there are probably at work other forces than the usual solvent power; but until the whole theory of solution is better understood we must be content to treat of these allied phenomena under the same general head of "solubilities." Any extended discussion of the methods ordinarily employed in determining solubilities, and the precautions necessary to insure accuracy, would perhaps hardly be in place here. Directions for making such experiments may be found in several chemical handbooks-for example, in Fresenius's "System of Instruction in Quantitative Chemical Analysis," or, better, in the original memoirs of those chemists who have occupied themselves with the experimental determinations of solubilities, references to which may be found in the body of this work. It may, nevertheless, be well to remark here that the text-books do not usually lay sufficient stress upon the preparation of the solution of the substance under examination, and yet this is the single fundamental point of a correct determination, the other steps of the process being altogether subsidiary, and, in general, easy of execution, as well as comparatively free from sources of error. It is commonly stated that an exactly saturated solution of a salt may be prepared either by exposing a large excess of the salt to the action of the solvent during several hours at the desired temperature (method by digestion), or by heating a mixture of the salt and solvent until a strong solution has been obtained at a temperature higher than that at which the determination is to be made, and then cooling this solution to the desired degree, and maintaining it at this point for some time in contact with crystals of the salt, the whole being frequently agitated (method of cooling).

CHEMICAL NEWS, July 2, 1864.

not always easily to be detected unless comparative solutions are prepared by the method of digestion, and the length of time required by any given solution to assume the normal condition is a point not readily ascertained. Gay-Lussac, in his classical memoir upon the solubility of salts in water, ‡ enjoins the necessity of maintaining the final temperature constant during at least two hours. His own experiments were made in the cellar of the Observatory at Paris, in which the thermometer varies but a fraction of a degree centigrade in the course of the year; they are unquestionably correct in themselves, and there can be little doubt that his statement regarding the preparation of normallysaturated solutions by the method of cooling is true, not only for the limited number of salts upon which he operated, but in general for crystalline substances. Yet the rule seems hardly safe to be followed in all cases by experimenters less favourably circumstanced, and it is obviously inapplicable to numerous uncrystallisable substances, or those liable to pass into an amorphous gumlike condition, or to undergo other molecular changes. The difficulty of avoiding supersaturation is, moreover, illustrated by the experience of Legrand, who found that solutions might become supersaturated to a certain extent even while they were actually boiling.||

Indeed, it is my opinion that, next to the impurity of the material operated upon, by which many published determinations have unquestionably been vitiated, there is no source of error so grave, none which has so seldom been fully guarded against, or so often altogether overlooked, as this tendency to supersaturation.

On the other hand, in the preparation of solutions by the method of digestion a difficulty is encountered in the tendency of many substances-like arsenious acid, for example-to dissolve with extreme slowness. This can, however, be overcome by the exercise of patience, and, in any event, admits of being detected and controlled. It would, therefore, appear that, where practicable, the method by digestion should generally be preferred, at least for temperatures low enough to insure the experiments against the influence of evaporation. The completion of the solution can then always be ascertained by determining from time to time the amount of substance dissolved, the operation being considered finished when the results of two of these tests accord with each other. As frequent agitation is indispensable, some process of stirring by machinery moved by clockwork, analogous to that described in Mohr's Lehrbuch der Pharmaceutischen Technik, might probably here be used with advantage. Kemp's regulator for maintaining constant temperatures might also be found serviceable in some cases.

Annales de Chimie et de Physique, 1819 (2), 11, 298.

"Dans chaque cas il faut maintenir constante la temperature finale saline, pour être bien assuré de sa parfaite saturation."

"Il semble d'abord que pour avoir cette température, il n'y a que observer celle à laquelle le sel commence à se déposer; mais on n'aurait ainsi rien de constant, il faut prendre celle qui a lieu pendant que le sel se dépose. En effet, j'ai remarqué que la dissolution pouvait se saturer malgré le mouvement d'ebullition, et attendre une température de plus en plus élevée; mais aussitôt que le sel se dépose, lo thermomètre redescend en un point ou il se tient parfaitement fixe," -Ann. Ch. et Phys. (2), 59, 428.

Compare Berzelius in his Lehrbuch, 3, 32, et seq. **Liebig and Kopp's Jahresbericht, 3620, 10,612, 12,709; also Journal of the Franklin Institute, (3), 25,319.

Royal Institution. The general monthly meeting will be held on Monday, July 4, at two o'clock.

PHYSICAL SCIENCE.

Suggestions for a Thermo-Spectrometer, by WILLIAM CROOKES, F.R.S.

WHEN soda is brought into a flame and the emergent light examined in the spectroscope it is seen to be of one degree of refrangibility only. Over the whole range of the visible spectrum, from the lowest red to the extreme violet, only one narrow line vibrates in synchrony with the heated soda atoms. But Professor Magnus has lately shown that heat is likewise given off by the soda flame in very large quantities, and these heat rays are just as much a portion of the spectrum as the visible or the actinic rays. If, therefore, we had some means of observing the thermic spectrum we should find that the sodium spectrum was far from simple. Optically, sodium stands next to thallium in the simplicity of its spectrum; at ordinary flame temperatures, thallium radiates one, whilst sodium radiates two homogeneous rays; these latter so close together, as to appear but one in ordinary instruments, Other metals give more complicated visible spectra; the activity imparted to them by the high temperature is expended over a wide extent of the coloured spectrum, whilst with sodium and thallium, the energy is concentrated into one line; hence the wonderful luminosity of these homogeneous lights, and the excessive brilliancy which they exhibit in comparison with other coloured flames given by different metals. Sodium, however, spends some of its vibrative energy in the invisible heat spectrum, and it would be a matter of considerable interest to examine whether the other metals, such as lithium and thallium, which give simple optical spectra, also radiate heat rays. This has indeed been found to be the case with lithium, which is stated by Professor Magnus to act like sodium, and it is probable that thallium would act in a similar manner; but from these theoretical considerations we should not expect to find that metals which yielded complicated optical spectra, such as barium or copper, would also emit heat rays of great intensity. Physicists now require a thermospectroscope, or, rather, spectrometer-an instrument which will enable them to examine and map out the thermal lines of the spectrum, as accurately as this can be done with the visible and photographic portion. With light (or rather with heat radiations) of tolerable intensity, this would not be difficult to effect. A single row of antimony and tellurium bars, soldered together at their alternate ends, as in the ordinary antimony-bismuth thermo-electric pile, could be securely cemented to a solid plate of glass and ground perfectly flat. This flat side could then be cemented to a permanent support of glass, porcelain, ebonite, or other suitable non-conducting material, and the other side (after removal of the temporary glass support) should be likewise ground down, until the series of bars was no thicker than a card. This side should now be cemented on to the same kind of supporting material as was used for the other side, and the whole firmly and securely sealed up at the sides, so as to leave only the ends exposed. The end of this pile would now be in the form of a very narrow line, which might be half an inch or so in length, and would consist of the extremities of ten or a dozen couples of antimony and tellurium bars, each not larger than a pin. The extremities of this battery being connected with a very sensitive galvanometer; the pile, upon being carried along the ultra-red end of the spectrum, would instantly reveal when a ray of heat shone upon it, by a deflection of the needle; and the comparative intensities of the

in the Chair.

Thursday, June 16, 1864. Professor A. W. WILLIAMSON, Ph.D., F.R.S., President, THE minutes of the preceding meeting having been read and confirmed and the several donations to the Society's library announced, Charles Tomlinson, Esq., Lecturer on Science in King's College, London, was formally admitted a Fellow of the Society, and the undermentioned gentlemen were duly elected by ballot-viz., G. W. Knox, B. Sc., Bristol; Michael Carpael, New Bond Street; Sidney Pontifex, Leadenhall Street.

A paper, by C. SCHORLEMMER, Esq., "On the Identity of Methyl and Hydride of Ethyl," was read by the Secretary. The author commenced by referring to an experiment in which equal volumes of chlorine and the gaseous hydride of ethyl were exposed to diffused daylight, and the chlorine substitution products thus obtained were collected and purified in the manner fully described in a communication to the Royal Society, entitled "On the Action of Chlorine upon Methyl." A colourless mobile liquid, condensed in the receiver, which consisted chiefly of chloride of ethyl, C,HCl, boiling at 11° C. Besides this compound, a small quantity of a liquid having a higher boiling-point was formed, from which monochlorinated chloride of ethyl, C2HCl, could be isolated. The results of these experiments differed widely from those of older researches of Frankland and Kolbe, who studied the action of chlorine upon the gas obtained by treating cyanide of ethyl with potassium, which they first considered as methyl, but afterwards recognised as hydride of ethyl. By the action of one volume of chlorine upon one volume of hydride of ethyl, they obtained one volume of hydrochloric acid and one volume of a gas having the composition CH,Cl, which substance was believed not to be chloride of ethyl because it could not be condensed at -18° C. Frankland showed afterwards that by the action of two volumes of chlorine upon one volume of hydride of ethyl, a liquid substitution product was formed, but that by acting with one or two volumes of chlorine upon one volume of methyl only, gaseous chlorine compounds were formed. and Kolbe performed their experiments with perfectly dry gases, whereas the author employed them in the moist state. It therefore became necessary to repeat the experiments of Frankland and Kolbe exactly under the circumstances described by them, only on a larger scale; but the author obtained results coinciding with those already described as having been furnished by the moist gases. One volume of chlorine and one of methyl, or one volume of hydride of ethyl, in the dry or moist state, yielded one volume of hydrochloric acid and chloride of ethyl, which always contained a small quantity of its chlorine substitution products. From these experiments it would appear that there does not exist any chemical difference between methyl and hydride of ethyl, and that the lowest isomeric terms of the two series of radicals and hydride stand in the same relation as the higher terms do. With author considered that, like all isomeric hydrocarbons of regard to differences in their physical properties, the these groups, they exhibit a close agreement, and only in their degrees of solubility in water was a slight disagreement noticed. The values of the coefficients of absorption of methyl are a little smaller than those of hydride of ethyl, but the curves representing these coefficients are

Frankland