March 21, 1889J 



NA TURE 



503 



fluoride, by Prof. T. E. Thorpe, F.R.S., and Mr. F. J. Hambly. 

 Gaseous hydrogen fluoride, on being heated from a few degrees 

 above the boiling-point of the liquid, shows a rapid decrease in 

 density, owing to the dissociation of Hj^Fr molecules ultimately 

 into HF molecules, the course of the dissociation being similar 

 to that observed in the case of nitrogen peroxide and acetic acid. 

 The density of the gas at about 32° corresponds with that required 

 for a molecule HoFo, but a careful study of the molecular break- 

 ing down of the vapour as it is eflfected by changes of temperature 

 and pressure shows that there is no evidence for the existence of 

 such a molecule. At a temperature of 26° •4, the lowest tem- 

 perature observed, the density of the gas corresponds with a 

 molecular weight of 5 1 '2 (H3F3 = 60), and from this point the 

 process of dissociation is perfectly continuous until the tempera- 

 ture increases to about 60°, when the density corresponds with 

 that of a vapour consisting wholly of HF molecules. In the dis- 

 cussion which followed the reading of the paper, Prof. Ramsay 

 said that Prof. Thorpe, in speaking of the analogy of the results 

 obtained in the case of hydrogen fluoride with those of the 

 brothers Natanson for nitric peroxide, had pointed out that these 

 latter afforded insufficient proof of the higher 'limiting value of « 

 in the formula N,,©,,, ; and that this limiting value was also un- 

 known in the case of acetic acid, of which the vapour-density also 

 increased with fall of temperature. Now there were three separate 

 lines of argument leading to a knowledge of the higher limiting 

 formulae of these bodies which had been pointed out by himself 

 and Dr. Young, and of which the data were to be found in papers 

 published in the Philosophical Transactions, in the Philosophical 

 Magazine, and in the Chemical Society's Transactions. The first 

 of these has reference to the alteration of density of the saturated 

 vapour with fall of temperature and corresponding fall of pres- 

 sure. It is argued that the density of the vapour of a su1)stance 

 must necessarily, at any given temperature, be higher when the 

 vapour is on the point of condensation than when it is un- 

 saturated. Hence, if it can be proved that the density of the 

 saturated vapour of bodies like nitric peroxide and acetic acid 

 shows no signs of increasing beyond those required for the re- 

 spective formulae N2O4 and C4H8O4, such formulae must denote 

 the limit of complexity of the molecules, in the gaseous state at 

 least. To ascertain such a limit, Dr. Young and the speaker 

 constructed from the Natansons' data for the relations of volume, 

 pressure, and temperature of nitric peroxide, and their own data 

 for the vapour-pressure of that body, isothermal curves in which 

 pressures formed ordinates and vapour-densities abscissas. The 

 terminal points of such curves are characterized by rapid in- 

 crease of density without rise of pressure, and, in fact, denote 

 that the substance is no longer in the gaseous state, the vapour- 

 pressure of the liquid having been reached. The densities of 

 the saturated vapour therefore will correspond with the angles of 

 union of the isothermal curves with horizontal straight lines re- 

 presenting condensation to liquid under vapour-pressures con- 

 stant for each temperature. By joining with each other such 

 angles of union for each temperature a curve is obtained ex- 

 pressing the densities of the saturated vapour in relation to 

 pressure. It is evident from inspection of such acurve for acetic 

 acid, shown in a plate in the Transactions of the Chemical 

 Society, 1886, 806, that the line of zero-pressure would be cut at 

 the density 60, corresponding with the formula C4HSO4 ; a 

 similar curve can be constructed from the Natansons' results 

 and Ramsay and Young's determinations of the vapour-pressures 

 of nitric peroxide, and this intersects the line of zero-pressure at 

 a point corresponding with the vapour-density 92, equivalent 

 to the formula N.,04. The second argument is as follows : 

 Representing the relations of temperature and pressure of a 

 " perfect " gas for any given constant volume, p = c . t, where 

 r is a constant and t absolute temperature. This is the 

 equation to a straight line ; such a line is termed an isochoric line 

 or isochor ; its point of origin for a perfect gas is absolute zero of 

 pressure and temperature. If a different volume be chosen, the 

 slope of the line is different. Now it is clear that if a given 

 volume of gas contains 2« molecules, pressure will rise with rise 

 of temperature at twice the rate that it would if the given volume 

 of gas contained n molecules. Constructing for nitric peroxide 

 and for acetic acid, on the assumption that they are perfect 

 gases, diagrams showing the relations of pressure and tempera- 

 ture for the formulae NO, and C2H4O2 at such volumes that i 

 Ljramme occupies, say, 1000 c.c in each case, the resulting 

 straight lines will manifestly differ in slope from those corre- 

 sponding to the respective formulae N2O4 and C4H8O4, the 

 pressure in the latter case not rising so rapidly with rise of 



temperature owing to the smaller number of molecules in that 

 volume. But we know that the actual behaviour of these 

 bodies is not that of perfect gases. The line representing the 

 relations of pressure and temperature, should at high tempera- 

 tures, when the substances exist in the molecular states NOj and 

 C2H40._,, nearly coincide with the theoretical line for these 

 molecular states ; and at low pressures and temperatures with 

 the line denoting the molecular condition N2O4 and C4Hg04. 

 The data of actual experiment show that such is the case. The 

 S-shaped isochoric curve trends so that it is probable that it 

 would become tangential with that expressing the behaviour of 

 molecules of the higher formulae, showing no signs of cutting it 

 as it must needs do were still more complex molecules capable 

 of existence. The third line of argument is derived from the 

 application of Raoult's method to a solution of nitric peroxide 

 in acetic acid, and the results obtained show that the molecular 

 weight corresponds closely with the formula N.JO4. — Contribu- 

 tions to the chemistry of lignification : the constitution of the jute 

 fibre substance, by Messrs. C. F. Cross and E. J. Bevan. The 

 authors describe the results of a fuller study of the ligno- 

 celluloses (cf. Chem. Soc. Trans., 1882, 90; 1883, 18).— The 

 atomic weight of chromium, by Mr. S. G. Rawson. The 

 atomic weight of chromium, as determined by converting a 

 known weight of ammonium bichromate into chromic oxide, is 

 found to be 52"o6l. 



Linnean Society, March 7. — Mr. Carruthers, F.R.S., 

 President, in the chair. — Mr. J. E. Harting exhibited specimens 

 of a South American Bat {Noctilio leporinus) alleged to be of 

 piscivorous habits, and which, through the kindness of Sir 

 William Robinson, the Governor of Trinidad, had been for- 

 warded from that island by Prof. McCarthy, together with a 

 Report on the subject. From this Report, it appeared that the 

 stomach of one specimen, opened within half-an-hour after it 

 had been shot on the evening of December 29, "contained 

 much fish in a finely-divided and partially-digested state." In 

 three others procured at 6 a.m. the following morning, the 

 stomachs were empty. On the morning of December 31, at 

 3 a.m., numbers of these bats were observed returning to their 

 caves : two were shot, and "both contained considerable 

 quantities offish." Prof. McCarthy added that in the stomachs 

 of other specimens examined by him fish scales were undoubt- 

 edly present. Of the specimens forwarded in spirits to this 

 country, two had been skinned and the stomachs and intestines 

 examined by Mr. Harting. The sac-like stomach was much less 

 muscular than might be expected in a fish-eating mammal ; but 

 in one of them (the other being empty) fragments of a finely- 

 striated and iridescent substance resembling fish-scales were 

 found. A discussion followed, in which Prof. Howes and Mr. 

 W. P. Sladen took part, the conclusion being that, although 

 there was no a priori improbability in the alleged piscivorous 

 habits of this bat, it could hardly be accepted as a fact until the 

 fragments, supposed to be offish, were really proved to be so by 

 careful microscopical and chemical examination. — A paper was 

 then read by the Rev. Prof. Henslow on the vascular systems of 

 floral organs, and their importance in the interpretation of the 

 morphology of flowers. The author drew attention to the im- 

 portance of this class of observations, as supplementing develop- 

 ment and teratology ; for, by referring all organs back to their 

 " axial traces," their real origins could generally be discovered. 

 Taking the words metaphorically as "floral units," he explained 

 how they can, as it were, give rise to axes as well as to all kinds 

 of floral appendages. Quoting Van Tieghem's definitions of 

 axial and foliar characters, the former was shown to be subject 

 to exceptions. After describing the arrangements of the cords 

 in peduncles and pedicels, in which endogens often have the 

 cords as regularly placed as in exogens, the author explained the 

 different ways by which pedicels and umbels are formed in each 

 class respectively. The " chorism " and union of cords were 

 illustrated and the effects produced. Considerable light was 

 thrown upon the cohesion and adhesion of organs, and the 

 interpretation of the " receptacular tube" and "inferior 

 ovary " was shown to depend upon the undifferentiated state of 

 the organs when in congenital union. The true nature of axile 

 and free central placentas was revealed, so that in the case of 

 the former, \\ ith scarcely any exception, the axis takes no part 

 in the structure, all "carpophores," " stylopods," &c., being 

 simply the coherent and hypertrophied margins of carpels. 

 Similariy, the free central placenta of Primula received its inter- 

 pretation as consisting of the coherent and ovuliferous baser 



