Injiiioicc of External Conditions on Tricotyly. 455 



tricotyls has increased, it is plain tiiat the factors under- 

 lying- these differences are by no means simple. 



This is further shown by the fact that a repetition of 

 the experiments does not always produce the same result. 

 Thus in Ainaranttis speciosus (1897), and ScropJiularia 

 nodosa (1898), I was not able to observe any influence 

 exerted by the various external conditions on the propor- 

 tion of tricotyls, although the experiments were carried 

 out on a large scale. I also found that if I sowed sam- 

 ples from the same lot of seed in two successive years, 

 the mean ratio of tricotvls in the harvest of the two 

 cultures was the same, (e. g., Oenothera hirtella, 1898 

 and 1899). 



Moreover, unfavorable conditions can sometimes, by 

 diminishing the yield, lead to an increase in the per- 

 centage values. This is shown by an experiment which 

 I made in 1898 with Antirrhimnn ma jus. I covered half 

 of my culture with a cage made of fine metal gauze 

 painted black, after the plants had developed a stem of 

 10 centimeters above the cotyledons. Within the cage 

 it was considerably darker than outside, and the plants 

 grew very wxakly, produced only few branches, and de- 

 veloped but a small number of flowers and fruits on the 

 terminal spikes. Fertilization was artificial, and all the 

 plants were guarded from the visits of insects. The 

 yield In the cage was 0.5 cc. per plant, and outside 1 cc. 

 per plant. In the former case as many fruits as possible 

 were gathered but in the latter the upper flowers of the 

 spike were not fertilized. The material used in this 

 experiment consisted exclusively of the offspring of a 

 single parent of 1897 which had had a value of 14%, 

 but tricotylous and dicotylous seedlings were both used, 

 each group being planted half within the cage and half 



