45 c. c. of a solution saturated at 9°.7 left 0.1266 at 150°. Accordingly, 1 part of the anhydrous salt dissolves in 359-3 parts of water grm. residue 1 part of the hydrous salt dissolves in at 90.7. 309-3 parts of water Another preparation* of the barytic salt, after three crystallisations, was treated in the same manner. 0.3398 grms. gave 0.1456 grm. barytic sulphate. Barium .... Found. 25.19 (C,H,NO,) Ba+ 4H2O. The determinations of solubility were made on 15-30 c. c. of The above salt crystallises in forms much resembling those assumed by a a and y nitrobenzoate. The component crystals, however, of each group are much thicker and are terminated by less acute angles than in the other two cases; the groups themselves, also, are of relatively much smaller size. The fusion-point of the hydric salt was determined on a specimen which had been thrice crystallised and was quite white. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 140-8 140 1 140-3 140-7 140 9 1415 140·9 140·9 140 4 141.3 142 141 141 150 160 160 170 170 The first fusion-point has been confirmed by other determinations. The amount of crude substance originally obtained precludes the serious influence of any traces of the γ modification which might be supposed present. (5). Action of heat on a Hydric Nitrobenzoate and on Barytic Nitrobenzoates.-a hydric nitrobenzoate, prepared from the barytic salt at the fifth crystallisation, was heated to 137°, and maintained at that temperature for a few minutes; it had fused *For this specimen I am indebted to my friend Mr. John Ferguson. VOL. XIX. 2 D long previously. After rapid cooling it was converted into barytic salt, and the solubility of this was determined. 30 c. c. of a solution saturated at 14°5 left 0-0913 grm. residue at 150°. Accordingly, 1 part of the anhydrous salt dissolves in 328.5 parts of water 1 part of the hydrous salt dissolves in 286.9 parts of water at 14°.5. y and 8 barytic nitrobenzoates change somewhat on heating to 150°. Residues from various determinations of the solubility of these salts were dissolved in water and crystallised; the motherliquids contained bodies of greater solubility than the original nitrobenzoates, especially in the latter case. (y) 50 c. c. of a solution saturated at 19°.7 left 0.1296 grm. residue at 150°. Solubility (of supposed anhydrous salt) 1 in 385 1 parts of water. (8) 15 c. c. of a solution saturated at 15° left 0·0784 grm. residue at 150°. Solubility (of anhydrous salt) 1 in 191-3 parts. A yellow coloration accompanies this change. On the other hand, a barytic nitrobenzoate shows scarcely a sign of alteration under these circumstances. Observations. The experiments recorded in (1), (3), and (4), taken in connection with those described in my previous paper, show that there are bodies having the composition of hydric nitrobenzoate which melt at temperatures ranging from 128° to 141°. The disputed number (127°) given by Mulder is therefore confirmed, although the substance with which he worked was not free from hydric benzoate. The fusion-point of these bodies evidently rises with the amount of force expended in their preparation; and their barytic derivatives decompose more easily the more preliminary work has been done upon them. When hydric nitrate is made to act upon hydric benzoate, the first nitrobenzoate obtained melts at about 128°. But, even here, if the action be considerably prolonged, the fusion-point may rise (as I had already shown) to 137.5°. Hydric benzoate (or the nitro-benzoate obtained by the first process) when treated with the ordinary nitro-sulphuric mixture for a short time, is transformed into a hydric nitro-benzoate fusing at 136°. Lastly, if this reaction be prolonged in its turn, a third nitrobenzoate is produced, melting at 141°.-Heat alone acts in a similar manner. It is shown in (1) that the fusion-point of the a-modification is much higher after the substance has been heated, and decidedly approximates towards that of the S-modification. The same is true of the y variety (3). Moreover the a nitrobenzoate, after a short and gentle fusion, yields a barytic salt which, in its solubility, strikingly accords with the d-salt. Naumann has made some careful experiments on the fusionpoint of hydric nitro-benzoate, which have led him to doubt Mulder's result. The bodies which yielded the number he mentions (141°) had, however, been submitted to repeated heating and a train of processes. A high fusion-point is precisely what might have been expected. It will be noticed that the pure barytic salts (a and 8) are even unable to retain a constant solubility on repeated crystallisation; the y-salt, on the other hand, exhibits a surprisingly perfect stability under the circumstances. Such transformations under the influence of heat are already numerous and familiar. Those which are now recorded appear to resemble most closely, in the suddenness and minuteness of the effect, the corresponding changes observed by Duffy in certain fats. I will now compare the nitrobenzoates I have discussed. a. The hydric salt melts first at about 128°. It yields a barytic derivative, which is but imperfectly stable on repeated crystallisation, less liable to decompose on heating than the a and y barytic salts, more soluble than the y salt, less soluble than the & salt. 7. The hydric salt melts first at about 136°. It yields a barytic derivative which is perfectly stable on repeated crystallisation, more liable to decompose on heating than the a-, but less so than the 8-salt; and less soluble than the a- or S-salt. 8. The hydric salt melts first at 141°. It yields a barytic derivative which is imperfectly stable on repeated crystallisation; more liable to decompose on heating than the a- or y-salt; more soluble than those two bodies; and also different from them in that its crystalline groups are of smaller absolute size, and the individual crystals have less acute angles. * Ann. Ch. Pharm. cxxxiii, 206. + Chem. Soc. Qu. J. v. 179. These differences find a parallel in the case of a- and B-lutidine described by Mr. Greville Williams*; and those bases also exhibit the striking resemblances which obtain among the above hydric nitrobenzoates.-ẞ hydric nitrobenzoate is already well known as a definite species; it differs by the broadest characters from those which precede. I will only remark that it appears likely from (2) that the substance obtained by Abel, and that described by Zinin and Sokoloff are identical with what I had examined under the name B-nitrobenzoic acid," and which had been termed "nitrodiacylic acid " by other authors. Finally, it will appear that the hydrogen in hydric benzoate is in such a condition that it is most minutely susceptible to the action of reagents. Three hydric chlorobenzoates are known,† in addition to the derivatives to which I have paid attention. We may legitimately suppose the most mobile of elements in the free state to preserve its character, to a certain extent, when combined; and in this respect, hydric benzoate only repeats on a more extensive scale what I have shown to be true for marshgas.‡ My best thanks are due to Professor Williamson for the use of his laboratory. XXXIX. Contributions to the Notation of Organic and Inorganic Compounds. By E. FRANKLAND, F.R.S. THE typical formulæ generally used to represent the different great families of organic compounds are far from fulfilling their functions in an equally perfect manner. Of the types in general use, those employed in the formulation of the artificial organic bases and the organo-metallic bodies are alone satisfactory; on the other hand, the water, hydrochloric acid, and hydrogen types fail more or less completely to give an intelligent representation of the atomic arrangement of the bodies which are formulated upon them. The cause of this comparative success in the one case and failure *Proc. Roy. Soc. 1864, 303. Ann. Ch. Pharm. cxxxiii, 241. Chem. Soc. J. [2] ii, 163. in the other is doubtless to be found in the almost exclusively synthetical origin of the alkaloïdal and organo-metallic compounds. The number and nature of the constituent radicals of these bodies became thus invested with a degree of certainty which, until recently, was almost entirely wanting in other families of organic compounds. The rapid progress that has of late been made in the application of synthetical methods to the production of most of the other great organic families, has rendered it possible to apply, with some prospect of success, a mode of notation, the principles of which were suggested in a lecture which I delivered at the Royal Institution in the year 1858.* This method was founded upon the carbonic acid, or, as it is now sometimes called, the marsh-gas type. The adoption of this type was, however, nothing more than the application to the compounds of the tetrad carbon atom, of the principles which I had previously employed in the notation of organo-metallic bodies containing metals of various degrees of atomicity. As examples of the application of this mode of notation to some of the chief organic families, the following may be adduced: : The meaning of these formulæ is so obvious as scarcely to require description; that of ethylic alcohol, for instance, represents an atom of tetrad carbon united first to the carbon of methyl, secondly to each of two atoms of hydrogen; and lastly, to oxygen, the latter dyad element being likewise united to an atom of hydrogen, which is thus linked, as it were, to the first carbon atom without being actually united with it. This is a true representation of the internal arrangement of the atoms composing alcohol, not, indeed, of their relative positions with regard to each other in space, but of the mode in which they are held together. Thus of the six atoms of hydrogen in alcohol, two only are united with the atom of carbon which, in the formula, is placed to the left of * Proceedings of the Royal Institution, vol. ii, page 540. + Phil. Trans. for 1852, page 441. |
