108 MISSOURI BOTANICAL GARDEN, 



thrown out by the glandular cells of the endostome, then 

 exerts a superior attraction perhaps quantitatively only, or, 

 it may be, qualitatively, and this leads the pollen tube to 

 enter the exostome, which is its normal course. I have else- 

 where dealt with the problem of the direction of the pollen 

 tube (Lloyd, 1902) suggesting that the pollen tube is guided 

 in its course by the differential distribution of a stimulant 

 arising from the egg-apparatus, chiefly the synergidae, and 

 that the whole phenomenon is chiefly a chemical one. The 

 evidence here before us prompts a modification of the above 

 suggestion, which however does no violence to its funda- 

 mental feature, to the effect that the stimulant may be 

 handled in a system of relays, the pollen tubes being guided 

 from one relay to the other. In the date the relay stations, 

 so to speak, are in the carpellary guiding tissue, the glandular 

 endostome tissue and the egg-apparatus and egg cell itself. 

 There is no objection to the assumption that these offer 

 either a renewed stimulus of the same kind, or even a dif- 

 ferent kind of stimulus each time, in view of the work of 

 Lidforss, who showed that positive curvatures are shown by 

 pollen tubes toward nineteen proteins of various groups. 1 ' 

 It appears not improbable that refined methods may discover 

 that where different guiding tissues occur, each one involved 

 produces its specific secretion which restimulates the pollen 

 tube from time to time on its course. 



The innermost layer of cells of the inner integument, 

 already mentioned, also presents a special degree of activity. 

 The cells become deeply columnar and distended at their 

 free extremities where they touch the embryo-sac. Thi 

 layer of cells is clearly a tapetum, and is analogous to that 

 described for the Compositae 3 and for a number of other 

 plants by various later authors. 



This tapetum is contributed to, to some extent, by the 

 adjacent ehalazal cells, so that it extends some distance down 

 the embryo-sac, and further down on the funicular aspect, 

 where it reaches as far as the antipodal region. Here it has 

 a distinctly pronounced development, 



* Lidforss, 1909. 3 Goldfluss, 1898-9. 





s 



