The pure and dry acid was dissolved without heat in some concentrated sulphuric acid, and the solution allowed to stand. Gradually the liquid became coloured, and simultaneously bands appeared crossing the spectrum of the transmitted light. The maximum of clearness and intensity was reached after four hours' standing at the ordinary temperature, and continued for several days. The spectrum observed is shown at Fig. 2; it is marked by a very dark band on D, sharp at its least refrangible side, but shading off into the green. Another band on F can be seen, but it is not well marked, since the rest of the spectrum is almost wholly cut off beyond this point. In addition to the bands and shading, strong red fluorescence was observed. This colouring matter likewise gives the madder tints with mordants; but these compounds do not present any characters worthy of special notice. When the spectrum of Anderson's alizarin is compared with that of the true madder alizarin, under like conditions (see Fig. 1), it will be seen how thoroughly distinct they are in optical properties, though so nearly identical in their chemical relations. Rufigallic Acid.-The striking parallelism in many physical and chemical properties known to exist between madder alizarin and rufigallic or parellagic acid rendered the comparison of the absorption spectra of the two bodies, under similar conditions, a matter of considerable interest. Rufigallic acid was prepared by digesting a solution of gallic acid in five times its weight of concentrated sulphuric acid at a gentle heat for a sufficient time. The purple-red liquid was then slowly poured into a considerable quantity of cold water. A brown flocculent deposit was thus produced; this was collected, washed with boiling water, and dried. Repeated resolution in sulphuric acid and precipitation by water rendered this pure. With alkaline carbonates and alum, dull purple liquids were obtained. By careful sublimation the powder may be made to yield beautiful orange-yellow crystals, very strongly resembling alizarin of madder prepared by a similar process. As in the latter case, a large portion of carbonaceous residue likewise remains behind. When the light transmitted by a sufficiently dilute solution of rufigallic acid in oil of vitriol is examined with a prism, the very characteristic spectrum represented at Fig. 3 is observed. The two least refrangible bands are those most strongly marked when the sun is the source of light; while the remaining two, of higher refrangibility, are only easily observed with artificial light. It is curious to remark here, that if a very weak sulphuric solution of the rufigallic acid be heated sufficiently, the absorption bands disappear, but become again visible as the liquid cools. The solution of the acid in carbonate of sodium is reddish brown, and simply cuts off the most refrangible portion of the spectrum, commencing at E; with ammonia a dirty purple-red liquid, which weakens that portion of the spectrum more refrangible than F. Alum gives with rufigallic acid a solution of a fine purple colour, which absorbs much light between D and F, leaving both ends of the spectrum clear. These reactions are remarkably different from those afforded by madder alizarin. A similar series of experiments was made with the crystalline sublimate obtained from the amorphous rufigallic acid, but the results were identical with those just described; we may therefore state that rufigallic acid sublimes unchanged. Murexides. The uric acid and caffeine inurexides are so closely alied in chemical constitution, as well as in many of their physical properties, that it appeared to be a matter of interest to compare the action of the magnificently coloured aqueous solutions of each body on transmitted light. Uric acid murexide was prepared by the action of ammonia or a mixture of alloxan and alloxantin, and the caffeine derivative by treating amalic acid with ammonia. The prismatic analysis of the light trans mitted by the aqueous solution of each did not show any marked difference in their optical properties. In both cases the spectrum was found to be much weakened in the green, but no bands were observed. On the addition of a little hydrate of potassium, the solution of uric acid murexide changed to a violet tint, but its action on the spectrum was not very perceptibly altered. A mineral acid discharged the colour from the original liquid. The solution of the caffeine compound, on treatment with hydrate of potassium, is simply weakened in colour, but it does not absorb light differently. The presence of a free acid does not diminish the intensity or alter the tint perceptibly. It was found, in prosecuting these experiments, that when either murexide was prepared by a process different from those above-mentioned, that product exerted the same action on the spectrum, but the relations to reagents were materially altered. We therefore learn that the two murexides remarkably resemble each other in optical as well as in general characters; so that, when prepared by similar methods, they do not differ from each other in any important point more than will two samples of the same murexide, when both have been produced by different processes. I may now mention a curious result obtained when applying the murexide test in a peculiar way to a solution known to contain pure caffeine. The liquid was acidulated with hydrochloric acid, warmed, and then a crystal of chlorate of potassium dropped in; after digesting for a sufficient time, the whole was allowed to cool, and some strong solution of ammonia carefully poured into the test-tube, so that the column of specifically lighter liquid might float on the test solution. This was then left at rest for twenty-four hours, at the end of which time the two liquids were found to have partially mingled, and produced a purple-red solution, which on prismatic examination was observed to cut off two sharp bands in the spectrum, as represented in Fig. 4. May we not infer from the results of these experiments that uric acid and caffeine are each capable of yielding several different coloured products ? ous solutions of the colouring matter other chemical agents react in nearly the same way as in the aqueous decoction. Pure colourless hæmatoxylin (C.HO3), when allowed to oxidise by contact with the air affords a solution which reacts in precisely the same way as the aqueous infusion of the dyewood. Brazil Wood. The aqueous infusion of the wood of the Lima variety of Casalpina crispa* is of a pale brownish yellow colour, and acts in a remarkable manner on the light which it transmits. Fig. 6 represents the absorption which it produces in the spectrum. The first or least refrangible band is strongly marked, but the second is lighter, and rendered less evident by the considerable weakening of the spectrum beyond it. The addition of a dilute acid but little alters the co!our of the solution; but, on examining with the prism, it is now seen to produce but one ill-defined band, intermediate in situation between D and E. On the addition of a small quantity of caustic ammonia to the infusion, the liquid changes to a fine ruby-red colour; it then absorbs the single band on the confines of the green, as represented in Fig. 7, while both ends of the spectrum are perfectly bright and clear. When alum is warmed with the aqueous infusion, the characteristic bands before apparent are replaced by a general absorption of the green rays. The alcoholic tincture of Brazil wood affords precisely the same reactions. Chevreul was, I believe, the first to separate a red crystalline substance from Brazil wood; to this body he gave the name Brazilin, as he supposed it to exist ready formed in the tree. Preisser, however, some years later, showed that colourless crystals can be obtained from the alcoholic extract by agitating it with hydrate of lead and subsequently decomposing the resulting compound with hydrosulphuric acid. The colourless substance so prepared he named Brazilin (C.H14O3), and Chevreul's red compound Brazilein (C18H28O). Notwithstanding the doubts entertained of the accuracy of Preisser's statements, I must say that I have succeeded in preparing a nearly colourless substance from Brazil wood by Preisser's process, which, when dissolved and treated with ammonia with exposure to the air, gave a coloured solution which acted upon transmitted light in a similar manner to the ordinary infusion of the wood. Of course the corroboration of Preisser's results in this direction in no way affects the objections which have been taken to his formula by Watts and others. Logwood. The light transmitted by the aqueous decoction of the wood of Hematoxylum Campechianum when analysed by the prism gives the peculiar spectrum represented at Fig. 5. This shows a strongly marked absorption band but little more refrangible than the solar line D, the side next to the fixed line being sharply defined, but the second edge shades off gradually to near E. A very marked diminution of light is observed in the higher portions of the spectrum, commencing about F; the intervening green space is also somewhat obscured. With carbonate of ammonium the solution is of purple-red colour, but the band beyond D remains unchanged; a new band, however, Camwood, or Barwood.-The absorption spectrum appears, rather faintly marked, midway between b and afforded by the alcoholic solution or infusion of camF; the remaining portions of the spectrum are much wood is very remarkable on account of the consideraless weakened than in the last case. When the soluble refrangibility of the bands which characterise it. tion of the colouring matter is warmed with alum, the band near p is alone apparent, the rest of the spectrum being nearly free from shade, with the exception of a slight absorption, extending from the least refrangible edge of the band to midway between c and p; this is the point from which all light is observed to be cut off in the very strong solution. These are the most striking reactions of the aqueous decoction. The alcoholic tincture of logwood is pale yellow, and simply absorbs the blue rays from about b, leaving the remainder of the spectrum unchanged. If to the alcoholic liquid a drop of caustic ammonia be added, a magnificent purple-red colour is produced; this solution gives the single absorption bands just beyond D very strongly marked, and leaves both ends of the spectrum perfectly clear and bright. In weak spiritu Fig. 8 represents the appearance which we observe on examining the light transmitted by such a solution with the prism. It will be remarked that, in addition to the bands situated in the green and blue spaces, a slight shading on a is perceptible. If to the alcoholic liquid we add a drop of strong ammonia, the solution becomes of a dull purple-red colour, but gives the peculiar spectrum shown at Fig. 9. The first and least absorption occurs a little beyond D; the second band is strongly marked, but its borders are ill-defined. Potash acts in a similar manner. When boiled with the addition of a single drop of strong alum solution, the two bands of the simple alcoholic tincture are destroyed, and replaced by a shading of the spectrum Pernambuco wood gives similar reactions. between E and F. If to the original liquid a few drops of dilute nitric acid be added, a similar result is obtained; but in this case the absorption takes place between D and E. low the actual boiling point of a portion of the oil, are entirely wanting, and their absence unavoidably greatly modifies the resulting product. If, then, we wish to distil a small quantity of hydrocarbons analogous to the manner in which this process is carried on on a large scale, we have to employ the process of "crack ON THE DISTILLATION OF HYDROCARBONS. ing only. But even then very rarely has the time BY JOSEPH HIRSH, PH.D. On reading the note of Prof. B. Silliman, on his "Examination of some California tar,"* it occurred to me that similar experiments, when carried on with a view of ascertaining the qualitative and quantitative result of illuminating oil, should be executed under circumstances more parallel to the distillation on a large scale, in order to arrive at an approximately true result. During distilation of natural hydrocarbons on a commercial scale, i.e. in large stills, the process of "cracking" always takes place to some degree; or, in other words, as all native hydrocarbons consist of mixtures of those substances of greatly varying boiling points, all hydrocarbons of high boiling points contained in such mixtures are during distillation exposed to various degrees of temperature below their own boiling point, as long as those hydrocarbons of lesser gravity and lower boiling point have not been removed by distillation. It is this exposure to a lower degree of heat than corresponds to the distilling point of an oil of definite gravity which comprises the operation of " cracking." If we consider that the stills used ordinarily in this country vary in capacity from 500 gallons to 20,000, and in single instances, as in the refinery of Reese and Graff, Pittsburgh, Pa., to 40,000 gallons a piece, the discharge of which varies in duration from one to six days, it will be evident that frequently, as in the last case mentioned, a portion of the oil, which may enter the still with a boiling point of, for instance, 400° F., will remain exposed for more than 100 hours to various degrees of increasing heat, all of which are below its own boiling point, as the heat of even the strongest fire is absorbed by the evaporation of the more volatile portions of hydrocarbons. of executing such a distillation of a few gallons been extended beyond a small number of hours, and the consequence is a distillate with a comparatively high boiling point. The difference between this last named distillation and that on a large scale is the same as the one between distilling coal for the production of illuminating gas and that for producing coal oils. The former, producing tar of great specific gravity, is, for the reasons mentioned, even on a large scale, carried on in comparatively small low retorts, while for the production of coal oils, the larger revolving retorts are found more desirable. In these, the oily vapours are exposed to a cooler temperature than their own, with every revolution of the retort, and are in this manner br. ken up into oils of lighter gravity, distinguished from the tar oils by their fitness for illumination. If, then, a few gallons of hydrocarbons are distilled with a view of presenting the probable results on a large scale, we ought to consider the amount of heating surface, which ought to be small, and only at the bottom of the still, together with the arrangements made for cooling the sides and upper portions of the distilling apparatus, as th's materially influences the process of "cracking." This may be more comprehensible by noticing the rules which by experience I found to regulate the d ́stillation of carbon oils on an extensive scale. 1st. The difference of temperature between the actual boiling point of oil of definite gravity, and of the temperature to which it is raised, is proportionate to the effect of the process of "cracking," i. e., the more the temperature of the actual boiling point of oil of definite gravity is above the temperature to which the same oil is raised, the greater is the quantity of light oil obtained. If then we wish to reduce the gravity of a very heavy oil greatly, we shall have to employ an exceedingly low temperature, so low sometimes as to suspend actual distillation for a short time. 2nd. The gravity of the distillate, resulting from re During this normal state of distillation in a still of such dimensions, the process of cracking" will be taken advantage of in a supreme degree without any especial efforts or loss of time, the resulting distillate being of a light specific gravity and corresponding col-duction of temperature, will be directly proportionate our, while only a small portion disti's over as paraffin oil, the latter being due to over heating, which with a small quantity can hardly be avoided in so large a still, on account of the peculiar shape which has to be given to it to insure sufficient strength, and the effect upon that of the fire. To this point I shall return again. If we compare the distillation of 5 or 10 gallons with the above we find that in applying fire freely, the entire quantity may be distilled off in less than an hour, and cannot likely take more than a few hours-referring, as before, to a moderately intense heat. Here, according to the construction of the still or the manner of applying heat, the distillate will contain hydrocarbons of as high a specific gravity respectively, and as high a boiling point, as the crude oil did before entering the still, or even of a higher one, if the oily vapours were over heated. In this case the two leading items in the distillation on a commercial scale, viz., time and a temperature be * CHEMICAL NEWS, No. 436, p. 171 (Am. Repr., June, 1868, page 257). to said reduction, i. e., if we reduce the temperature to a degree at which only naphtha of 0.700 boils, the resulting distillate will possess a specific gravity of 0.700 regardless of the gravity of the original oil. This law enables us to produce a distillate of any desired gravity (below that of the oil before distillation) from any crude oil, and a due regard to it enables us to produce an illuminating oil of the same specific gravity without great quantitative loss from the light Pennsylvanian oils, as that produced from the heavier Ohio, Canada, or California oils. In distillation the temperature therefore should always be reduced to the boiling point of oil of the specific gravity desired. 3rd. The difference between the temperatures of the two boiling-points, viz., of the oil being subjected to distillation, and of the desired distillate, is in direct proportion to the height of the still employed, or, which produces the same effect, to the facility for cooling the upper portions of the still. According to this law a heavy oil will yield the readier a light distillate the higher its vapours have to rise before leaving the still, because the reduction of temperature in those higher portions of a retort which are more remote from the source of heat acts upon the vapour of the oil in fine division, and reduces their gravity more readily than the compact liquid oil. If the heat is applied solely to the bottom of the still, while its sides and top may be exposed to a current of cool air, the reduction of temperature of the oil vapours takes place similarly to, and can be controlled better than, the cooling in high stills without this provision. I would mention that I deem his concluding assertion "The transformation of light oils into denser products like tar, to result not as has been supposed by some, from the addition of oxygen producing an oxidised body, but by the removal of successive atoms of hydrogen in the form of water,"-simply a different mode of expressing the opinion of others referred to, as the removal of hydrogen in the shape of water is certainly a perfect and true process of oxidation. Western Chemical and Starch Works, 201 and 203 South Water Street, Chicago, June 24th, 1868. ON A NEW ALKALOID CONTAINED IN The arrangements for cooling the sides and top of a still must therefore be the more complete the lower or smaller the still employed is. This law also teaches us that stills to be used for "cracking" oil should have a flat bottom, and should have the flues arranged in such COMMERCIAL ANILINE AND ISOMERIC WITH a manner as to permit the restriction of the fire to the bottom only, which is necessary to the process of "cracking." Where superheated steam is used as the heating medium it ought to be applied at the bottom only. Where the dimensions of a still become as huge as was mentioned in the beginning of this article, a flat bottom would hardly be strong enough; in this case the boiler shape is usually employed. In order to have a practically sufficiently large heating surface, the fire here has to reach up to a certain distance on the round boiler. If the quantity of oil present in the still is so small as to be below the boiler surface exposed to the fire, the raising oil vapours will be superheated on this surface, and the resulting distillate will be of greater specific gravity, and of darker colour than it normally would have been, the product resembling more the oils resulting from the distillation of coal-tar. For this reason the residuum in such stills, after reduction to the quantity mentioned, is frequently removed to smaller stills. This diminution of temperature of the oil vapours causes a partial condensation and redistillation of the oil, which diminishes the colour and gravity of the oil. 4th. The intensity of the process of "cracking" is proportionate to the suddenness with which the oil vapours are condensed before leaving the still. The thorough application of this principle produces more rapidly those results mentioned as necessary in the preceding paragraph. 5th. The difference in gravity between that of the oil distilled and of the desired distillate is in direct proportion to the quantity of water produced in the process. If from a very heavy oil an exceedingly light distillate is to be produced, the proportion of water is so immense as to show occasionally a distillate of water with but a minute percentage of oil floating on its surface. In all such cases small black particles of carbon float on the top of the water, forming an intermediate layer between the latter and the oil. The quantity of carbon separated in this manner is also proportionate to the quantity of water distilled over resp., to the intensity of the process of "cracking" employed. This carbon is mechanically carried over by particles of water, which in contact with oil always produce violent ebullition. These laws are the same with hydrocarbons distilled under the ordinary atmospheric pressure as with those distilled in a vacuum or under increased pressure. In the two last-named cases the variation of boiling point corresponding to these different degrees of pressure has to be taken into consideration. Referring once more to the article of Prof. Silliman, TOLUIDINE. BY M. ROSENSTIEHL. By the transformation of the toluol of coal tar into toluidine a non-crystallisable alkaloid is obtained. This has been remarked by several observers-amongst others by MM. Coupier and Graefinghoff; the latter explains the fact by admitting that toluidine may exist under two modifications, one liquid, the other solid, but he points out no other distinctive characteristics. Similar differences have been observed in nitro-toluol; M. Jaworsky obtained this product in the crystallised state; and M. Alexeyeff demonstrated that reduction transformed it into toluidine, totally and immediately crystalline. M. Kekulé concludes from these facts that the liquid nitro-toluol hitherto known must be impure. About three years ago M. Coupier introduced to commerce a toluidine but partially crystalisable, and containing only 20 per cent of aniline. To this liquid toluidine appertains the valuable quality of forming, with arsenical acid, a rel colouring matter analogous to fuchsine. This property has hitherto been attributed to an admixture of aniline and toluidine; and the liquid form of M. Coupier's product was supposed to be due to the presence of a certain quantity of aniline. Several facts are opposed to this view of the question. There is the constant boiling point of the alkaloid (198), then its elementary composition, which answers to the formula of toluidine; and, lastly, the circumstance that the yield of pure red dye is vastly super.or to that obtained from admixtures of pure aniline and crystallised toluidine in the most favourable proportions.-(Bulletin de la Société Industrielle de Mulhouse, p. 264, 1866.) When liquid toluidine is cooled to below zero, and a drop of water thrown in, part of the mass solidifies, and the crystallised toluidine separates. The liquid which remains still possesses the boiling point and the composition of toluidine; but when treated with arsenic acid the yield of red is less-15 per cent. in lieu of 45 per cent. This liquid is the best primary substance for the preparation of the new alkaloid. It is to be transformed into oxalate, which is exhausted by ether free from alcohol; the undissolved portion consists of pure oxalate of toluidine: the part dissolved consists of an oxalate, crystallisable from ether, alcohol, and water. When decomposed by soda, this oxalate furnishes a liquid alkaloid. To ascertain that this was indeed a simple alkaloid, it was transformed into a chloride and separated by successive crystallisations into four deposits; each of these deposits was recrystal ised, and same conditions produces a blue black. I ought to observe, in concluding this summary, that the purity of the alkaloids used by me in the course of these experiments was verified by the laborious but sure process employed upon the pseudotoluidine.-(Compt.s Rendus, lxvii., 45.) its solubility in water determined. The constancy of this solubility was looked upon as a proof of complete purity. By the aid of this salt the free alkaloid employed in the following experiments was prepared. Recently rectified over fused potash, this base is colourless, but colours gradually in the air. It remains liquid at -20, its odour resembles that of toluidine, its density is 10002, and it boils at 198 at a pressure of 744 millimetres. Analysis led to the formula C,H,N. ON THE PROXIMATE ANALYSIS OF COALS. This formula was controlled by the analysis of its oxamide, which crystallises in beautiful silky looking needles, easily obtained quite pure. This amide differs from the oxalate by a molecule of water only: { ((C,H)2) CO2O2-H2ON, CO HO (C,H10N)2 S The alkaloid regenerated from this amide was also analysed, and the results accord with the preceding analyses. As it is, therefore, an isomer of toluidine, will call it pseudo-toluidine, that being the only name I feel justified in giving it, while still ignorant of its exact constitution. It is sufficiently plentiful-the liquid toluidine of M. Coupier contains about 36 per cent, commercial aniline often more than 20 per cent. Pseudotoluidine does not seem to be identical, either with methylaniline, which boils at 192 (Hofmann, 1850, Annalen der Chimie und Pharmacie, lxxiv., p. 150); or with the alkaloid described by Limpricht, under the name of benzylamine (loc. cit., cxiv., p. 304), which boils at 183°. I have carefully compared the crys'alline form, and the solubility of the chlorhydrate and oxalate, with those of the corresponding salts of aniline and toluidine. The establishment of this difference is very important. The form of the chlorhydrate is undeterminable; the salt of pseudotoluidine forms an orthorhombic prism; that formed by the toluidine is clinorhombic. Their solubility in the same order are respectively, in a hundred parts of water, 129 at 17, 7; 37, 5 at 15, 5; 22, 9 at 11°. The coloured reactions of this solution are alike different to those of aniline and toluidine; but I shall return to this part of the subject in another communi cation. BY PROFESSOR GUSTAVUS HINRICHS, CHEMIST OF THE GEOLOGICAL SURVEY OF IOWA.* COAL is not a simple chemcial combination, expressible by a chemical formula; but it is the final residuum of vegetable matter having been exposed to a long-continued and complex process of addition and subtraction. An elementary analysis will therefore not teach us much in regard to the nature of the combustible; for who would dare to make any conclusion concerning the peculiar combination of the elements thus determined? Even the heating effect calculated from this elementary analysis is not more trustworthy than the valuation by the reduction of lead. The proximate analysis, on the contrary, enables us to learn something in regard to the real nature of the fuel. The moisture and the ashes are both not only diluents of the fuel, but in themselves obstacles to the effectiveness of the same; the vaporisation of the moisture causes a serious loss of heat, while the ashes, by hindering a complete combustion, and by the heat they contain when dropped through the grate, constitute another loss. By furthermore determining the total amount of volatile matter, we learn both the percentage of coke in the fuel, and the amount of carbon (fixed combustible) and bitumen (volatile combustible matter). Although neither of these two products can be considered as simple chemical compounds, it is nevertheless of the utmost practical importance to know these two quantities, because of the great importance of coke and gas in the arts. The yield in gas of a fuel is no doubt measured by the percentage of bitumen, at least for coals from the same basin-coals which therefore may be supposed to have passed through nearly the same chemical history. When heated with arsenic acid, pseudotoluidine In taking charge of the chemical labours of the does not produce red, but if combined with pure crys- Geological Survey of Iowa, I had grave doubts in retallised toluidine, it yields an abundant red colouring gard to the value of this proximate analysis of coals. matter, containing at least 50 per cent of a rosaniline No investigation as to its accuracy, nor as to the best salt; during the reaction much aniline also distils. An method of conducting the work, had come to my explanation of this remarkable fact would require a knowledge. The European chemists seem a most exprofound study of the composition of the new alkaloid. clusively to rely on the elementary analysis, while in Here is another circumstance not less striking. When the great Government surveys of this country the mixed with aniline and treated by arsenic acid, pseudo-proximate analysis seems to have been almost as extoluidine produces plenty of red colouring matter, similiar to fuchsine, but differing from salts of rosaniline inclusively practised. But while the former may readily be turned into approximate determinations of the heatthe solubility of its base in ether, and the still greater ing effects of the fuel, the latter have, to my knowsolubility of its chloride in water. It appears from its ledge, never been used for such purposes, nor was it at origin to be an isomer of rosaniline. all apparent that they ever could be thus made useful. In regard to the first condition of all quantitative determinations, that of giving constant results for the same substance, but few observations were accessible, and these rather increased my first distrust. Thus Whitney, nowhere in his report on the coals of Iowa (Geology of Iowa, vol. i., part 1), gives any data whatever in regard to this most important question, although he devotes a very large space to the subject. The anilines now employed in the manufacture of fuchsines contain much less aniline than formerly, but I have proved that they contain as much more pseudotoluidine. At the same time the composition of the reds is changed; the red which I have described must now be recognised, in addition to the chloride of rosanilin whose formula Hofmann has so clearly estab lished. Salts of pseudotoluidine, mixed with chlorate of copper and applied on cotton, give a fine black; this black approaches violet, whilst pure aniline under the From the American Journal of Mining, published in advance of the official report. |