112 FRANCIS E. LLOYD 



has further become clear to me from my studies of changes in 

 ripening fruits (and of analogous changes in other parts) that the 

 hydrolysis of the interwalls (usually regarded as middle lamel- 

 lae or primary membranes) depends on the secretion of appro- 

 priate enzymes by each particular cell, and indeed by each par- 

 ticular contingent region of the protoplast. In consequence, two 

 adjacent cells may be set free of all other surrounding cells, but 

 themselves remain attached to each other. The softening and 

 the deliquescence of the persimmon fruit may be prevented by 

 killing the tissues just as may, also, the process leading to non- 

 astringency^ so that the early death of any of the component 

 cells of the mesocarp for any cause, spontaneous or otherwise, 

 should lead to the same result coextensively with the cells or 

 tissues involved. This, in point of fact, actually occurs in the 

 red-coloured parts of seeded fruits. Not only do the red tannm 

 cells themselves remain attached to each other, for many days 

 after the fruit is completely non-astringent, but the adjacent 

 parenchyma cells also remain attached to them, while the un- 

 coloured tannin cells become separate from each other (but not 

 from adjacent red tannin cells) and from the parenchyma cells. 

 To prove this it is necessary only to shake up portions of over- 

 ripe mesocarp in water, when the cells will either fall apart or 

 cling together in accord with the conditions just described. 



There remains the question as to the reason why the presence 

 of seeds should influence the rate of the process of becoming non- 

 astringent and the consequent death of many tannin-cells, to- 

 gether with the resulting reddening by oxidation of their tannin. 

 Any answer must however be of a general enough nature to in- 

 clude the fact that the same result occurs, although less exten- 

 sively, yet independently of seed production, and the further 

 fact that the occurrence of seeds in certain varieties does not 

 induce the changes described. We have also the observation 

 that, in the variety Taher 23, the entire fruit had become non- 

 astringent even while still green (at Auburn, Alabama, in 1912), 

 Whatever the answer, the fact to be accounted for is the abnor- 



^ Vinson, A. E. The stimulation of premature ripening by chemical means. 

 Journ. Amer. Chem. Soc, 32: 208, February, 1910. 



