i I 



676 



REPORT — 1884. 



and ammonium carbonate, which cannot exist at all in the pfaseous state, the- 

 temperatures of volatilisatiim do not form a curve. When the disaociation was 

 considerable but not complete, aa in the case of phtbalic and succinic acids, an 

 indication of a curve was observed at low pressures, but it differed widely both in 

 form and position from that representin"' the vajiour pressures or pressures oi 

 <lissociation. As tlie dissociation decreases the curves approach each other inoiv 

 closely, and they appear to bo coincident in ihe case of ammonium cldoride and 

 nitric peroxide witiun the limits of temperature at which observations were luade, 

 and at which the amount of dissociation is ])robabIy small. With acetic acid 

 very numerous observations proved the ])erfect coiucidence of the curves. 



The results appear to be unfavourable to the view that when liquefaction of n 

 gaseous stable substance takes place, gaseous molecules coalesce to form mor.; 

 complex groups of molecules, and that these complex molecules dissociate wheu tLe 

 substance is vaporised. 



b. On Molecular Volumes. Bi/ Professor Wili.ia:m Ramsay, Ph.V. 



]. The object of this research was to asjertaiu whether, as has been long taken 

 for granted, the boiling-points of compounds under equal pressures really afibvd 

 suitable points for a comparison of their molecular volumes. The expeiinient.i 

 described in detail in the original paper were made during the years 1880 and ISSl, 

 and a preliminary notice was read to the Chemical Society in the spring oi' tlia: 

 year. An account of experiments by W. Staedfl appeared subsequently, in wliioli 

 he showed that tiie element chlorine possesses at least three dilferent volunu'3 in 

 combination. Subsequent researches by Lossen, Schilf, and others have since tiiat 

 lime tlioroiighly proved that no element enters into combination with invarialjie 

 atomic volume. The experiments made by the author decisively prove that in tlio 

 following series of compounds : water, methyl alcohol, ethyl alcohol, propyl 

 alcohol, isopropyl alcohol, isobutyl alcohol, and ether; the value of tlie group (JIL 

 is by no means constant, while at the builing-points of the liquids at low pressuroi 

 the value is approximately constant, fluctuating between 17'5 and 22 ; at liigli 

 temperatures the diiference becomes nuicli more apparent, attaining, at pressures of 

 20,000 mins. (wliicli was the highest measured), the greatest irregularity. Tiius 

 the difference between the molecular volumes of ether and isobutyl alcohol, two 

 isomeric substances, amounts to a total of about 20 units, whereas the hypothesis 

 that at the boiling-points under equal pressures the molecular volumes are com- 

 parable, aud for isomeric substances should be equal, no difference between thes^: 

 substances sliould be observable. 



2. It was supposed that as liquids at high temperatures corresponding to lii;:)i 

 pressures are extremely com])ressible, the volumes might be comparable, ])rovideJ 

 they were compared under critical pressure ; the temperatures, however, at wliicii 

 they were compared .still being those at wliich their vapours exert equal pre.-^sure. 

 To elucidate this point, careful measurements of the compressibility of these liipiids 

 •were made, and also of their critical temperatures and pressures, and it was found 

 that when the liquids were at the temperatures corresponding to equal vapoiir 

 pressure, b>it exposed to their critical pressures, no correspondence between tlieir 

 molecular volumes was observable. 



3. As a last alternative it was thought possible that if the liquids still at 

 temperatures corresponding to equal vapour pressures, coiUd have existed under 

 no pressure, some basis of comparison might be found. Necessarily such a state is 

 unrealisable in pructice, but as the compressibility of the liquids had been deter- 

 mined, it was calculable. Again, it was found that in this hypothetical comlition, 

 although tlie relative volumes at high temperatures were considerably altered, ye'o 

 no point of com])iirison had been reached. 



4. The autlior therefm-e concludes that, contrary to what has usually heen 

 supposed, tlie boiling-points of liquids under whatever pressure they may be taken 

 arc not suitald ! femjieratures at which to comjiare their molecular volumes. 



Still it cannot Ije denied that a certain regularity is noticeable. Tlio 

 approximate constancy of the atomic volumes of elements which is made by Kopp 



A 



