PERMEABILITY AND THE PROTOPLASMIC MEMBRANE 283 



weighted at one end and floating in a lake. The weighted end is 

 down, owing to greater attraction at that end (due to gravity in 

 the case of the logs; solubility or like factor, in the case of the 

 molecules). Such logs or molecules in vertical formation and 

 therefore closely crowded, or in oblique and entangled formation 

 and therefore separated, will form a surface layer more or less 

 permeable to other bodies, depending on the size of the latter 

 (see further, page 289). 



Estimations of the size of colloidal pores have been made for a 

 number of nonliving substances. Bjerrum gives 1 to 90 m/x 

 as the size of colloidal membrane pores. The smallest silica gel 

 pores are 5 m/i. 



The sieve hypothesis has its weaknesses, but it has often been 

 criticized on unjustifiable grounds. Thus, the behavior of the 

 two forms of rf-galactose and other sugar-permeability phenomena 

 already referred to cannot be explained in terms of size of the 

 sugar molecules. The same is true of glycerol and monoacetin, 

 the molecules of which are of very nearly the same size, yet they 

 enter the cell (as judged by hemolysis of red blood corpuscles) 

 at a rate nearly fifteen times as fast in the one case as in the other 

 (220 sec. for glycerol and 15 sec. for monoacetin). The first of 

 these is very poorly soluble in lipoid material, and the second a 

 good lipoid solvent; solubility, therefore, rather than size of 

 molecule, is the more likely determining factor. But actually, 

 as in the case of most if not all sugar-permeability phenomena, 

 the question is not one of permeability primarily but of metab- 

 olism. The cell is freely permeable to both forms of the sugars, 

 but it uses one in metabolic processes and not the other. The 

 greater "permeability" of the one is therefore a matter of relative 

 concentration within and without the cell. 



In an attempt to make the sieve hypothesis fit certain selective 

 properties of the protoplasmic membrane, the following specula- 

 tion has been indulged in. The different rate of penetration of 

 substances of the same molecular dimensions is explained by 

 imagining that the cell membrane is a sieve with pores just large 

 enough to permit the substances, say ions, to enter. Once inside, 

 the ion is supposedly converted from the ionic condition, in 

 which it entered, into a molecule, or from a smaller molecule into a 

 larger one, and so cannot now leave through the sieve ; other ions 

 are not so changed and may therefore pass out again. There is 



