126 MISSOURI BOTANICAL GARDEN 



rise to any of the abundant sugar, one of the alternative 

 views advanced (Bigelow, Gore, and Howard, 1906). 



Ovule. Stage I. At the time of pollination there is 

 tannin distributed throughout the whole of the ovule, as 

 . shown by a diffuse reaction. There is, however, a much 

 greater quantity in the embryo-sac, inner integument, chal- 

 aza, raphe and pedicel (f. 2). The endocarpal pollen-tube 

 guiding tissue shows a similar amount of tannin. 



Regarding, for a moment, the anatomical facts recited 

 above, it is difficult to avoid the conclusion that there is 

 special nutritive activity connected with tannin botli at the 

 antipodal end of the embryo-sac and in the tapetum. In 

 the Eleagnaceae (Servettaz, /. c), while tannoid substances 

 occur in the nucellar cells beneath the antipodal cells, and 

 do not occur in the embryo-sac, it is argued that the anti- 

 podel cells play an active role. The conception of the physio- 

 logical importance of these cells, in spite of their frequently 

 small size and lack of histological peculiarities, is one to 

 which we have become accustomed through the work of 



Westermaier, Goldfluss, Lloyd, and others, 17 but it still re- 



mains to determine in detail what their method of operation 

 may be. Servettaz' suggestion should prove fruitful. 



In the date, although the antipodal cells are small, their 

 anatomical relations are peculiar. In the definitive embryo- 

 sac, they are placed at the end of a slender pit, surrounded 

 by elongated cells with thickened walls (f. 4). The thicken- 

 ing appears to be due to swelling, and this must, I think, 

 be of significance since it is only the walls forming the anti- 



podal pit which are swollen (f. 1, 3). Whether this is the 

 same sort of thing observed by Servettaz (7. c, p. 354) or not, 

 [ cannot say. He appears to regard the thickening and 

 gelatinization of the antipodal cell walls observed by him 

 as an accompaniment of disintegration. In the case before 

 us, the swelling of the cell walls is not due to digestion, since 

 their persistence (f. 5) shows the contrary. They appear, 

 however, to be full of tannin, as thev show a definite reaction 



17 Coulter and Chamberlain, Morphology of Angiosperms. 



