288 [Dec. I, 



It is well known that when a bladder nearly full of alcohol is 

 immersed in a vessel of water, the water will pass through the 

 bladder and become mixed with the alcohol, though little or none 

 of the alcohol will pass out to become mixed with the water. This 

 result constitutes a particular case of the phenomena described by 

 Dutrochet under the terms Endosmosis and Exosmosis. Ever since 

 the attention of men of science was called to these phenomena by the 

 French philosopher, they have attracted a large share of attention ; 

 and many chemists and physiologists have at different times en- 

 deavoured to define the laws that regulate these actions, which were 

 regarded by many as proceeding from a force that had not pre- 

 viously been recognized. 



It was not, however, till Mr. Graham adopted the simple expedient 

 of placing the two liquids under experiment in contact with each 

 other without the intervention of any septum, that a distinct idea of 

 the steps of the operation was obtained. Mr. Graham proved that 

 these movements were due partly to the action of the liquids on each 

 other, and partly to their action on the septum. In three papers 

 published in the ' Philosophical Transactions ' for 1850 and 1851, he 

 traced the laws of "Liquid Diffusion," and was thus enabled to 

 show that the process in liquids is regulated by principles closely 

 analogous to those which in earlier researches he had demonstrated 

 to prevail in the diffusion of gases. 



In these experiments a number of small jars of about 4 oz. capacity 

 were prepared with necks ground to a uniform aperture of 1 inch 

 in diameter. Into these jars the trial solutions were poured : each 

 jar was then closed by a glass plate, and placed in a cylindrical vessel 

 containing about 20 oz. of distilled water ; the mouth being sub- 

 merged at least one inch below the level of the water. The glass 

 plate was then cautiously removed. Each solution having been thus 

 treated, the jars were left for several days undisturbed at a steady 

 temperature. After a sufficient length of time the mouth of each 

 jar was again closed with a plate of glass, and the jars were with- 

 drawn. The water contained in the outer vessel was then in each 

 case evaporated, and the salt that had passed into it determined by 

 weight. Saline substances were thus found to be divisible into cer- 

 tain groups of bodies of equal diffusibility ; the rates of diffusion of 

 the different groups being connected by simple numerical relations. 



