462 



SCIENCE. 



[N. S. Vol. IV. No. 92. 



which have an actual independent existence 

 entirely in abeyance. It would, doubtless, 

 be possible to give such definitions of atom 

 and molecule as shall include only the re- 

 sults of our actual empirical knowledge 

 and mathematical deductions therefrom. 

 Such definitions would stand on a basis 

 which is incontrovertible, and it is incon- 

 ceivable that even the overthrow of the 

 atomic theory, or any knowledge which 

 may be gained in the future as to the na- 

 ture of matter, could change them. 



That such definitions possess great value 

 is certain; and that it is very important to 

 distinguish sharply between our positive 

 knowledge and speculations and theories 

 based on that knowledge, every one will 

 admit; but, after. all, unless we join that 

 school of philosophy which teaches that 

 there is no real existence outside of our 

 own minds, there is some reality at the 

 basis of and behind all the phenomena 

 which we investigate. And it is the prov- 

 ince of science to find out the truth about 

 every real existence of which we can secure 

 any tangible evidence. Our knowledge of 

 atoms and molecules as actually existing 

 particles is, doubtless, a purely speculative 

 •deduction from a multitude of diverse phe- 

 nomena, and yet the mental picture con- 

 nected with the concept has been, and still 

 •continues to be, of very great value in the 

 development of our science. These mental 

 pictures are vague and in many respects in- 

 .complete, it is true, and they doubtless 

 'do not correspond closely to the real exist- 

 ences for which they are, at present, our 

 best expression; but, to one familiar with 

 the very practical results which have been 

 ^obtained in the domain of structural chem- 

 istry, it is difiicult to conceive how such re- 

 sults would have been possible without 

 their use. 



"While physical chemistry has made little 

 •or no apparent progress toward the goal 

 marked out by Du Bois-Reymond, its prac- 



tical achievements during the last ten or 

 fifteen years have been very considerable, 

 and it is to these practical achievements 

 that I wish to turn our attention. 



Whenever the subject of physical chem- 

 istry is mentioned our thoughts naturally 

 turn to the subject of solutions. !N'ot that 

 physical chemistry has to do only or chiefly 

 with solutions, for, as I understand it, 

 physical chemistry has to do with all of 

 those physical properties of matter which 

 can only be understood by taking into con- 

 sideration at the same time its composition; 

 but rather because some of the most val- 

 uable results of physico-chemical researches 

 have been made in this field, and because 

 diversity of views has made this the chief 

 recent battle-ground in chemistry. 



Up to about eleven years ago our knowl- 

 edge of solutions was almost entirely of an 

 empirical character. No great generaliza- 

 tion, similar to those which had long been 

 known for gases, had been discovered. 

 In 1885 J. H. van't Hoflf^ proposed his 

 theory of osmotic pressure. The em- 

 pirical basis for this theory lay in the 

 experiments of many different work- 

 ers, some of them made many years 

 before. Studies in osmose date from the 

 early years of this century and, indeed, 

 some experiments were made more than a 

 century ago. It was not, however, till 

 1867 that the discovery of true semi-perme- 

 able membranes was made. In that year 

 M. Traube^ showed that membranes may be 

 prepared which will readily allow the pas- 

 sage of water, but which are totally imper- 

 vious to certain substances in solution. 

 Ten years later W. Pfefler^ conceived the 

 idea of preparing such membranes in the 

 interior of a solid partition. By exposing a 

 cup of porous porcelain to the action of a 

 solution of copper sulphate on one side and 

 of potassium ferrocyanide on the other, a 

 precipitate is formed within the mass of the 

 porcelain which is permeable to water, but 



