46 Flowers and their Unbidden Guests. 



contrary, that would reach the flower by creeping up 

 from below, would not give themselves the trouble of 

 climbing over the upper margin of the perianth, and so 

 making their entrance by the projecting stigmas, but 

 would get at the nectar by the much shorter and, 

 seeing that their approach was from below, much 

 more convenient way, namely, through the iatervals 

 between the deeply-divided segments , of the perianth. 

 They would, therefore, never touch the stigma at all, 

 and thus the nectar would be sacrificed without the 

 advantage of allogamy being attained. And as 

 autogamy is an impossibility in this plant, owing to its 

 being dichogamous, and owing to the above-described 

 relative positions of its anthers and stigmas, the visits 

 of these small creeping insects would put a complete 

 stop to the formation of fruit. But such creeping 

 insects are prevented from all possible access to the 

 flowers, inasmuch as the Polygorvu/m ampMMum grows 

 in water, which encircles the stalks of its iiiflores- 

 cence. But what if the water were to run off, and 

 the plant be left ^n dry ground? Now it is very 

 remarkable that when this happens special means of 

 protection are developed, which did not exist while the 

 plant was growing in the water. An iunumerable quan- 

 tity of horizontally projecting trichomes ^ or glandular 



' [As the term " trichome " is as yet hardly in common use in 

 English text-books it may be as well to explain that it is a general 

 name given in the higher plants to all such outgrowths as originate 



