30 Tannin in the Fruit of the Pcrsimmon [Sept. 



swelling of the tannin-mass results in a closer crowding of the 

 protoplasm, in its disruption and in that of its vacnoles and in 

 the concomitant stretching of the cell-wall. It is obvious, of course, 

 that, no matter how close the tannin-mass crowds the protoplasm, 

 there must be a film of water between them; indeed this must repre- 

 sent also, at least in part, the watery contents of the broken vac- 

 uoles. As a matter of Observation, one may see the vacuole burst 

 the tonoplast and spread out within the protoplasm itself. But, if 

 we cause such a swelled tannin-mass to shrink, we cannot find 

 evidence of free tannin in the space thus formed between the tannin- 

 mass and the protoplasm. The natural conclusion appears to be 

 that the tannin-mass gives out free tannin only upon swelling, as 

 if the particles of free tannin were entrapped by the meshes of the 

 associated colloid when these, under pressure, are contracted; and 

 escaped when, the pressure being released, there is room enough 

 to move. 



There is at least one condition, however, which seems to nega- 

 tive the view that the free tannin may not be found just outside the 

 tannin-mass. I have found that the tannin-masses in the fruit, 

 already referred to (when exposed to acetic acid for three days 

 and thereby killed), are found, after a lapse of three months, to 

 swell and burst the cell wall in water. They have, however, passed 

 into a condition of rigidity greater than that usually found in cells 

 which burst, since the rupturing of the cell results from the tearing- 

 open of the tannin-mass (fig. 27), due to the circumstance, appar- 

 ently, that the inner part absorbs water more rapidly than the 

 outer. Now, if there were free tannin distributed throughout the 

 tannin-mass, we should expect to find it diffusing from a part or all 

 of the freshly exposed surface (f.s, fig. 27), as indeed happens 

 when such is the case. The fact is, however, that free tannin 

 escapes only from the edges of the fracture (g, fig. 27), and, as I 

 have repeatedly assured myself by close Observation, from the 

 fissure between the torn edge of the cell-wall and the tannin-mass. 

 If a mass of such cells be mounted, so that they may remain clothed 

 by the pectose adhering to them, and water added, the free tannin 

 may be expelled slowly and may form precipitation membranes 

 (fig. 21). If an unbroken cell be tested for tannin, however, the 



