Tannin in Persimmons. 13 



tannin, both in the unruptured and ruptured cells, and is prob- 

 ably caused by the access of oxygen. During normal advance of 

 ripening no colo'"~*ion of the tannin-masses occurs. Aso * 

 states the contrarv, with whom, however, Howard takes issue, 

 although he finds a few red tannin cells, and gives a photograph of 

 them. It is clear to my own mind that red or brown tannin-cells 

 are such as have early died within the tissue, and become colored 

 by chemical change. Their appearance in Howard's photograph 

 suggests this. I have not seen any colored tannin-cells in un- 

 treated fruits which did not have every appearance of being ab- 

 normal in this sense. 



Conclusions. 



The tannin in the tannin-cells of the persimmon fruit when 

 unripe, is not wholly in a watery, or cell-sap, solution. It is rather 

 associated in part with a carrier in a manner analogous to a 

 gelatin, albumen, or other colloid-tannin union. 



The insolubility of tannin, (itself a colloid) in the ripe 

 fruit, is due to its intimate and complete association with this 



Fig. 1. Tannin cell showing the distribution of depressions (pits and canals) in the sur- 

 face of the tannin-mass occupied by plasmatic structures and sap-vacuoles. (n) Nucleus. 

 Fig. 2. Detail of the surf ace depression occupied by nucleus and vacuole- Spindle-shaped 

 carotin bodies in the plasma. 



Figs. 3, 4, and S. Tannin cells which burst and tore in distilled water. Fruit a week after 

 full ripening. 



Fig. 6. Tannin cell from a fruit ripened in ethyl nitrite. The tannin-mass has become 

 obular in shrinking. 



Fig. 7. End of a cell in a ripened fruit, showing the plasmolysed protoplasm, (t) tannin- 

 mass; (pr) protoplasm: (c.w) cell wall. 



Fig. 8. A similar cell after partial swelling of the tannin-mass on being placed in water, 

 (v) sap-vacuole. The turgid condition is due wholly to the pressure exerted by the swelling 

 tannin-mass. 



Figs. 9 and 10. Tannin-mass after normal digestion, from a completely ripened fruit, 

 which gave, upon eating, no astringency whatever. The cell-wall and protoplasm have 

 been digested. The bursting occurred either during mastication or in the stomach. 

 Fig. 11. A cell which burst after being washed, dried and subsequently placed in water 

 again. 



Fig. 12. Tannin-cell after prolonged treatment with dilute nitric acid. The granular 

 mass is still present at (g). Behind it is the portion of the cell-wall pushed 

 out during rupturing, A non-tanniferous colloidal shell (c) remains within the cell-wall. 

 Fig. 13. Detail of such a cell, drawn with a camera lucida, showing the colloid shell, proto- 

 plasm and cell -wall, from left to right. Optical section. 



Fig. 14. Cell bursting in very weak ferric chloride, in which the granular matter may be seen 

 escaping (as indicated by the arrow points) and collecting to form the granular mass, which 

 usually accompanies the extruded tannin-mass. Below, a similar granular mass extruding 

 independently. 



♦Quoted by Howard, I have not seen the paper. 



