THE PERMEABILITY OF PROTOPLASM 103 



phasized the fact that several surfaces may be involved, 

 and prefers to speak of surfaces rather than membranes, 

 since a definite membrane is not essential for semi-perme- 

 ability. 



To investigate such surfaces the marine alga Griffithsia 

 was employed. Within its cell wall a thin layer of proto- 

 plasm surrounds a central vacuole, which expands or 

 contracts according as water is taken in or given out by 

 osmosis. Such cells diminish in volume when placed in 

 hypertonic sea-water, but regain their normal size when 

 returned to their original surroundings. If the cell be 

 placed in hypertonic ammonium chloride, instead of in 

 hypertonic sea-water, contraction takes place, but the 

 inner wall shrinks to a greater extent than does the outer. 

 Thus the space between the two surfaces, which is normally 

 very small, may increase until in places it is equal to one- 

 third of the length of the cell. Thus a distinction must 

 be drawn between the outer surface of the protoplasm, 

 the plasmatic membrane, and the inner surface, the vacuole 

 wall. The explanation of such behaviour may be that 

 the outer is more permeable to ammonium chloride than 

 the inner. The salt would therefore have a stronger 

 plasmolyzing action on the inner surface, since the more 

 readily a substance penetrates, the smaller, is its efficacy 

 as a plasmolyzing agent. The alternative interpretation 

 is that the permeability of the surfaces may be altered by 

 the ammonium chloride, and the contraction may be due 

 to " false plasmolysis," as certain appearances of plas- 

 molysis produced by injurious actions are termed. The 

 behaviour of Griffithsia could then be accounted for by 

 the supposition that the false plasmolysis of the inner 

 surface was greater than that of the outer. Osterhout 

 inclines to the opinion that both effects contribute to the 

 result, 



