316 Dr. J. H. Long on the 



amount, the portion diffused was known. The amounts diffused 

 from similar solutions of ten different substances are given by 

 Beilstein; but, owing to the fact that he failed to state the exact 

 dimensions of his diffusion-glass (he mentioned that it con- 

 tained about 5*5 cubic centims.), it is impossible to compare 

 these results with those of Graham. The determination of 

 Fick's diffusion-constant from Beilstein's work is, of course, 

 impossible. 



In a paper published in 1857, Simmler and Wild* have 

 shown that in certain cases it is possible to determine the dif- 

 fusion-constant from a condition of the solution not yet sta- 

 tionary, as they claim was the case in the experiments of their 

 predecessors. 



They also suggested several new methods of observation. 

 One of these, a so-called optical method, was some twenty 

 years later applied by Johannisjanzf ; and in the meanwhile 

 another optical property of certain solutions was made use of 

 by Hoppe-Seyler| and Voit§ in similar investigations. 



Hoppe-Seyler and Yoit experimented with substances pos- 

 sessing the property of rotatory polarization — a solution of 

 grape-sugar for instance. Such a solution was placed in a 

 glass vessel with parallel sides ; a layer of water was then 

 brought over this without mixing with it. The sugar, diffu- 

 sing into the water above, imparted to this latter the rotatory 

 power. By observing the amount of rotation at different 

 heights in the glass vessel, data were obtained from which the 

 diffused sugar at these points could be calculated. 



Johannisjanz used a hollow glass prism as diffusion-cell. 

 The lower half of this was filled with salt-solution, the upper 

 with water. As the salt diffused upwards the refracting-power 

 of the latter was altered; and upon this was based the method 

 of determining the amount of diffused salt. For the details 

 of these processes I refer to the original articles mentioned 

 above. Although these three investigations were evidently 

 conducted with great care, it was found, however, that the 

 results obtained by Hoppe-Seyler and Yoit did not agree with 

 each other, and that the value of k, as determined by Johan- 

 nisjanz for NaCl, was quite different from that obtained by 

 Fick. 



The cause of these discrepancies remained unknown until 

 very recently, when Stefan || showed that the optical methods 

 are based on a false assumption, and that they can give 

 as a consequence only false results. He shows that a liquid 



* Pogg. Ann. c. p. 217. t Wiedemann's Annalen, ii. p. 24. 



X Medicinisch-chemische TJntersuchungen : Berlin, 1876. 



§ Pogg. Ann. cxxx. p. 227. || Wiener Sitzungsber. Dec. 1878. 



