CHEMISTRY. 349 



of pore-diameter) which will permit the more rapidly moving cation 

 of the membrane to pass nearly through the wall of the cell from the 

 outside, while the slower anion is only just within the interior mouths 

 of the pores. 



It has been noted, also, that the diameter of the pores must not 

 fall below a certain minimum magnitude, smce otherwise the mem- 

 brane is deposited, not within the mouths of the pores, but more or 

 less loosely upon the interior wall of the cell, where it is easily ruptured 

 by pressure. It has been stated, further, that the allowable difference 

 between maximum and mmimum diameter of pore is not very great. 

 The effective area of the membrane obviously depends upon the 

 diameter of the pores in whose mouths it is deposited, and the rapidity 

 with which the pressure will attain its maximum within a cell depends 

 on the area of the membrane. If, therefore, the texture of the wall 

 exceeds a certain degree of fineness — even though the membrane may 

 be correctly located, i. e., wholly within the mouths, of the pores — 

 the cell may be too ''slow" for useful employment. 



The foregoing statements will serve to explain the nature of some 

 of the difficulties encountered in the manufacture of porous cells for 

 the measurement of osmotic pressure. The limiting conditions of 

 successful cell production were apprehended very early in the course 

 of our work. It nevertheless required four years of experimental 

 work to produce the first faultless porous cell. 



The means by which we have been accustomed to secure the required 

 texture of cell-wall have been briefly indicated in other reports. 

 Unfortunately they are too elaborate and require too much special 

 training for general use. The art of cell-making has therefore been 

 confined, up to the present time, to the laboratory in which it origi- 

 nated. It is obviously desirable, however, that the porous cells which 

 must be used in the measurement of osmotic pressure should become 

 a commercial article, in order that the manifestations of this force 

 may be more widely studied from the experimental side. With a 

 view to simplifying our methods and thereby making feasible the 

 manufacture of cells by others, we have for several years (whenever 

 practicable) reverted to the problems of cell-making, and during the 

 past year more seriously and persistently than at any previous time. 

 Some progress has been made, especially in the matters of baking 

 and glazing the cells, both of which operations have been hitherto 

 subject to much difficulty and uncertainty — the former because of 

 its effect in detennining the size of the pores, and the latter because 

 '^ crazing," due to difference m expansion coefficients between glaze 

 and biscuit, can not be tolerated in osmotic cells. 



Previously we have never quite succeeded in removing old and 

 disabled membranes so as to leave the cells in perfect condition for 

 the deposition of new ones, though the causes of the comparative 



