EXPERIMENTS WITH TOBACCO BLOSSOMS. 15 



both sides of the pistil and be seen to terminate in the suture just men- 

 tioned. That there is :i close relation between the pollination of one- 

 half of the stigma and the setting of seeds in the corresponding half 



of the ovary is certain, but no attempt was made to ascertain whether 

 pollen grains placed on one-half of the stigma are everable to fertilize 

 ovules in the opposite half of the ovary. The fact that pollen tubes 

 all descend in the delicate conductive tissue occupying the center of 

 the pistil might indicate that the tubes are free to grow to any por- 

 tion of the ovary, but stained sections show that they grow parallel 

 with one another and seem to follow the long, thin-walled cells of the 

 conductive tissue (PI. II, tig-. 3). In the ovary the tubes make many 

 short turns in various directions (PI. II, tig. 5). 



While flowers two days before opening fall as completely as any. 

 when their stigmas are entirely covered with pollen, they nevertheless 

 recover from a slight premature pollination better than younger flow- 

 ers. This can be noticed by contrasting experiments 44 and 46 with 

 experiment 47. 



As will be noticed in the tables, several substances were substituted 

 for pollen. This was done in an effort to rind if other substances could 

 injure the stigmas and cause the flowers to fall as in the case of prema- 

 ture pollination. Nothing besides pollen was found to have this effect. 

 Its damage goes deeper than the stigma. 



As a mere irritation of the stigma is thought to have a tendency to 

 cause some flowers to set fruit, it may not be amiss to view these 

 experiments with this in mind. Of 60 emasculated flowers that had 

 their stigmas covered with substances other than pollen, 14 set fruits, 

 while of the 20 that were emasculated but never pollinated 2 set fruits. 

 As a general thing the capsules resulting from the flowers that were 

 not pollinated, and likewise those resulting from flowers whose stig- 

 mas were covered with some substance other than pollen, contained 

 onl}' small compressed undeveloped seeds, but the two pods obtained 

 in experiment 52 by treating the fully receptive stigmas with magne- 

 sium sulphate contained some spherical seeds of almost full size which 

 looked like good seeds, but when cut into proved to be hollow spheres. 



By combining like experiments in which flowers of various ages were 

 plentifully pollinated with fresh pollen, we get the following percent- 

 ages of seed pods, of which it can be said that none resulting from the 

 pollination of flowers earlier than one da} r before opening contained 

 g-erminative seeds: 



Per cent. 

 From 20 flowers pollinated four days before opening, 1 seed pod set. 5 

 From 40 flowers pollinated three days before opening, 2 seed pods set . 5 

 From 20 flowers pollinated two days before opening, seed pods set. 



From 40 flowers pollinated one day before opening, 31 seed pods set. 77 

 From 20 flowers pollinated one-half day before opening, 19 seed pods 



set 95 



From 20 flowers pollinated when fully receptive, 19 seed pods set. . . 95 



