1897.] NATURAL SCIENCES OF PHILADELPHIA. 171 



tion that a single flower failed to mature seed. It must certainly 

 be held remarkable that in a single plant, bearing in round num- 

 bers over 30,000 flowers, every one should bear two seeds. 



It has been contended that though plants may generally self-fer- 

 tilize when the agents for cross-fertilization do not attend, they are 

 so arrauged as to cross-fertilize when the agent does appear. As 

 the Heliophytum flowers are freely visited at times on my grounds 

 by insects, and especially butterflies, there might be some strength 

 in the point. I can, however, testify by an almost daily observa- 

 tion of my plant through the season, that minute flowers are only 

 visited by insects when others are scarce. Though I have seen 

 them visiting the flowers for several successive days, there are 

 many days when they do not visit them, and none were noticed on 

 the former until the beginning of August. A careful watching of 

 the anthers shows, however, the extreme difficulty of effecting cross- 

 fertilization. The anthers form a cone over the stigma, and the 

 pollen sacs burst almost simultaneously with the unplaiting of the 

 corolla. When the flower is a few hours old the stigma protrudes 

 slightly through the anther-cap, and is visible under a lens through 

 the very small orifice of the corolla- tube. Even admitting that the 

 flower has not fertilized at this early stage, and that the tongue of a 

 butterfly might introduce foreign pollen to it under certain cir- 

 cumstances, it would rarely, in any case, occur. It is well known 

 that all insects soon discover the easiest method of doing their 

 work. In this case there are five openings between the tube of the 

 corolla and the bases of the filaments, offering a wholly unobstructed 

 course to the creature's tongue. It would have to use considerable 

 force to insert its tongue under the anthers pressing down on the 

 stigma. It is inconceivable that the flower can receive any aid to 

 cross-fertilization in this way. But we may grant that a cross-fer- 

 tilization will result in a plant better fitted for the struggle for life 

 than one self-fertilized, and that a small percentage might become 

 cross-fertilized. The question of numbers again forces itself upon 

 us. How many of the seeds of any plant get a chance to develop to 

 a plant again bearing seeds ? How many of the 65,280 seeds of this 

 plant will probably mature ; will come to be seed-bearing plants 

 next year? Only a small percentage, in any case, ever do. In this 

 case surely very few will, and of these how many would those result- 

 ing from a " chance cross " give? 



