152 MISSOUKI BOTANICAL GARDEN". 



where it dips inwardly. The place of juncture becomes 



more and more obvious with age till the completion of 



secondary cell-wall thickening, when it remains visible on 



account of the irregularly matched cells. This is a plane 



of weakness where, during the initial stage of germination, 



the endosperm plug is released. 



The constant advance of the definitive endosperm opposed 



to the growing embryo results in the reduction of the zone 



in question. I have already pointed out that the secondary 



divisions in the endosperm cells which characterize it are 



always parallel to the integuments. They are therefore at 



right angles to the axis of the embryo. If these are taken 



as the criterion of the zone, it is seen that it is at last reduced 



to a thin layer of cells (b' y f. 19, 20), which are able neither 



to take on the definitive character of endosperm cells, nor 



entirely to resist the influence of the embryo. This is 



shown by the partial thickening of the walls, and in the 



partially digested condition of the contents. The shearing 



caused by the opposed directions of growth is also to be seen 



in the oblique and more or less distorted forms of the cells 

 (f. 20, 21). 



Tannin in the zones of digestion. — It has been shown that 

 during the third period of development tannin figures prom- 

 inently in the endosperm in connection with the laying down 

 of reserve cellulose. It has been pointed out, however, that 

 the reaction is not confined to the cells alone which are active 

 in this regard, that, in a word, it occurs everywhere in the 

 endosperm, the statement being intended to include the 

 zones of digestion. For example, ethyl nitrite material thir- 

 teen weeks after pollination showed that the tannin in the 

 tension tissue and in the digestion zones proper is in marked 

 amount, as it is also in the embryo. This is true of younger 

 stages, and of more advanced ones until the condition is 

 reached represented by figure 18, when, along with the gen- 

 eral reduction of tannin throughout the endosperm, the 

 amount in the digestive zones is also reduced. This disap- 

 pearance is not synchronous with that in the embryo, in 

 which tannin is discoverable for a longer period. It must 



