EMBRYO, SEED AND CARPEL IN THE DATE. 



137 



reaction in any case, the tannin being completely confined 

 to the cells in which the secretion at first takes place. The 

 genesis of the tannin may be observed very exactly in the 

 second foliage leaf, at the age indicated. It first appears as 



minute granules which beco 



finally 



form 



mass 



transitory 



must be regarded, in the absence 



evidence to the contrary, as aplastic. 



OIL. 



1 appears in the carpel at any time. 

 9 endosperm oil appears for the first time after about 

 nth week following pollination, before the endosperm 

 m is obliterated (f. 8-10). There is a gradual in- 



form 



about 



of the cell walls begins in the endosperm. As already 



time 



nection with the thickening walls. For a short period 

 (f. 26, 26a) at about the time of this phase of development, 

 the oil content is reduced, during which time the seed reaches 

 its definitive form. The oil content then increases again till, 



mature seed, the endosperm 

 3 foods, reserve cellulose an 



According to du 



Sablon (1897) 



9% 



sperm in the resting seed. I find no evidence of tannin 

 within the resting endosperm save in certain cases due to 

 irregularity of development already noted. 



The disappearance of oil and the beginning of cellulose 

 accumulation are without much difficulty explained on the 

 same grounds upon which we may also explain the accu- 

 mulation of oil as the reserve cellulose approaches a definitive 

 condition. The relation of tannin to these is more obscure, 

 but there is sufficient evidence, already advanced, to regard 

 it as related in some way to the changes which go on during 

 the thickening of the cell walls. 



