ON THE SPECIFIC VOLUMES OF LIQUIDS. G3 



It must be admitted, however, that certain of these deductions are drawn 

 from experimental evidence, which, in the light of our present knowledge, can 

 hardly be deemed sufficiently comprehensive to permit of such broad genera- 

 lizations. For example, the examination of only four liquids can scarcely 

 afford adequate proof of the universality of the statement that members of 

 the same chemical family have identical specific volumes. The conclusion 

 with regard to isomerides was necessarily based on limited proof, for the 

 reason that the number of cases admitting of examination was limited. The 

 number of isomerides has now increased a hundredfold, and we have become 

 more precise in defining their character. The validity of the law should be 

 tested by an examination of well-chosen and typical isomerides, especially 

 among the hydrocarbons. Such an examination would not only afford mate- 

 rial for solving the primary question, but would incidentally serve to show 

 whether the specific volumes of the component elements, carbon and hydrogen, 

 are respectively invariable, as stated by Kopp, no matter how these elements 

 may be arranged in a compound with respect to each other. 



It has been shown by Professor Koscoe that vanadium is a member of the 

 phosphorus group of elements, and that the vanadium trichloride of Eerzelins 

 is in reality an oxychloride of the composition VOCI3, corresponding to the 

 phosphoryl trichloride (POCI3). As both these analogously constituted hquids 

 are readily obtained in a state of purity and boil at moderately high tempe- 

 ratures, it seemed desirable to determine their specific volumes with a view 

 of obtaining further evidence on the isosterism of members of the same che- 

 mical family. As the result of a series of carefully conducted observations 

 made on preparations of a high degree of purity, we find that the specific 

 volumes of phosphoryl trichloride and vanadyl trichloride are distinctly 

 difierent, the chloride with the higher molecidar weight having the greater 

 specific volume. We have thus been led to reopen the whole subject. Start- 

 ing with the observations on the question of the specific volumes of members 

 of the same chemical family, we find that the result foreshadowed in the case 

 of phosphorus and vanadium is a general one, viz. that in a series of analo- 

 gously constituted compounds belonging to the same chemical family, as, for 

 example, the trichlorides of phosphorus, arsenic, and antimony, and the tetra- 

 chloride of silicon, titanium, and tin, the specific volume increases with the 

 molecular weight. 



We have completed the experimental work connected with the determina- 

 tion of the rates of expansion, boiling-points, and specific gravities, which 

 data (together with the molecular weights) are required to fix the specific 

 volume, of the following liquids : — 



Br CCl, PCI3 



ICl CBrCL PC1,C,K0 



C.,H,Br, PBr^ 



C,;H,IC1 SiCl, POCI3 



C.,H,C1,, TiCl, POBrCl, 



CH3CHCI, SnCl, PSC13 



CH,C1, VOCI3 



CHlBr", T,., , . . ^ AsF, 



CHCI3- ^^^y^^yljCH^^ ASCI3 



CHBr, and Heptane I ^ ^^ g^^.^ 



The labour of reducing the observations, and more particularly of calcu- 

 lating the empirical formuloe for so large a number of liquids, is necessarily 

 somewhat heavy and tedious ; its completion has been unavoidably delayed 



