164 PROCEEDINGS OF ^HE ACADEMY OF [1897. 



which were considerably over two feet tall, while again one often 

 conies across flowering dwarfs barely six inches high. 



I collected, on the excursion above referred" to, 25 spikes for ex- 

 amination, and of these 21 were staminate and 4 pistillate. Of 

 the latter closer examination revealed that on 2 of them some few 

 stamens with ripe anther cells were to be found, the anthers being 

 well-filled with pollen, fig. 11, p. The anther cells had burst, 

 and there is no reason why the pollen, in such cases, should not 

 fertilize the ovules in pistils on the same plant, especially since small 

 insects are always found inside the spathe, which may serve to dis- 

 tribute the pollen. Possibly these few stamens may help to ensure 

 fertilization in case cross fertilization should fail, which latter 

 method, for aught I know, may be the usual one. The stigma is so 

 remarkable that I could not refrain from sketching it, fig. 12. It 

 is densely covered with enormous club-shaped hairs which are ex- 

 tremely like the glandular hairs I found producing the jelly-like 

 secretion in the fruit of Peltandra undulata. Here and there I 

 found a pollen grain on these hairs,/). 



Finally : From the numbers cited above, although they can- 

 not be taken to represent the ratio in which staminate and 

 pistillate spikes are to be generally found, it appears, never- 

 theless, that nature wastes a great deal of energy to secure the 

 formation of fruit by such an excessive production of pollen. We 

 may suppose that all these staminate flowers are produced to ensure 

 fertilization by insects. There occurs to me another point of view 

 to which I am at present somewhat inclined : Every one must ad- 

 mit that in nature there is an inherent tendency under proper con- 

 ditions to produce flowers even if there may be little chance finally 

 of seed formation, although the ulterior object of the flower is then 

 not accomplished. "We know that in the genus Ariscema the 

 " flowers are monoecious or by abortion dioecious." It is quite pos- 

 sible that this abortion may have a direct bearing on the method 

 of reproduction described above. I wish to emphasize the point 

 that while the plant obeys this flower forming inclination, the sup- 

 pression of fruit production on those plants, which by abortion 

 develop only staminate flowers, may serve to increase the tendency 

 to the formation of underground runners. At all events, if we at- 

 tempt to realize the great number of plants of Ariscema triphyllum 

 produced by buds we must admit that the plant with stam- 

 inate flowers only, is possibly not less prolific than the fruiting one, 



