268 Cross — Eiipatoriiun ageratoides and coelestinum. 



fruits in twenty-three days, no of which were black and 

 swollen and contained well-developed embryos. Eight cap- 

 itula, selected from a portion of a panicle protected Sep- 

 tember 3d, cross-pollinated by hand September 19th, put in 

 press September 27th, produced 108 fruits. Of these forty- 

 two were white, shrivelled and imperfect, and sixty-six 

 were black and promised to be fertile. Plants protected 

 September 3d, close pollinated by hand September 19th, 

 the product gathered September 27th, produced 106 fruits 

 from eight capitula, forty-three of which were black and 

 swollen, and when dissected showed well-developed em- 

 bryos, while sixty-three were pale and shrivelled, with im- 

 mature embryos. Plants protected September 3d, left 

 undisturbed till gathered on September 21st, produced 245 

 fruits from sixteen capitula, only ten of which appeared 

 well-developed and fully matured. In order to test the fer- 

 tility of the embryos, I selected seeds showing the best de- 

 velopment, from plants grown under the four conditions 

 mentioned in the tables, and planted them January 7th, in 

 a large germinating dish between pieces of white felted 

 flannel. On March 31st radicles were seen protruding 

 from fruits grown under these conditions. This continued 

 until the middle of May, when at least seventy-five per 

 cent, had germinated. The radicle in germinating pro- 

 truded from the side of the fruit— Fig. 11— just above the 

 point of attachment to the stalk 1 . Twenty of the seed- 

 lings were potted, and made vigorous growth. None of the 

 seeds from plants cross or close pollinated by hand, or seeds 

 from plants protected and left undisturbed, germinated ; 

 although subjected to the same conditions as those grown 

 naturally. Reviewing the above observations, it seems 

 clear that self-pollination in the species of Eupatorium is 

 very rare indeed ; and that even when it does occur, the re- 

 sulting fruits are of weak germinating capacity. When 

 close pollination by hand is effected, a slight increase in 

 the production of good fruits is obtained ; but when cov- 

 ered flowers are cross-pollinated the increase is very strik- 



