Physiology. 121 



principal stimulus, causing in these closure, in other species opening, of 

 the flower. The difference in effect is paralleled in positive and negative 

 heliotropism and positive and negative geotropism. The influence of 

 different colors of light has never been tried. Chemical stimuli were used 

 by Tassi (1884). He found that flowers of N. alba and other plants were 

 unable to perform their movements under the influence of anaesthetics, 

 and in -some plants the color changed. 



The large and showy flowers of waterlilies are in many ways designed 

 for the visits of insects. The colors are mostly conspicuous and attractive. 

 N. caerulea is comparatively insignificant ; and N. rubra makes very little 

 showing at night, but in the early morning it is magnificent. N. rudgeana, 

 according to Caspary (1878), is often cleistogamous. The possibility of 

 such a thing is suggested by its regular occurrence in Euryale ; and 

 N. amazonum, with its odd habits, may well be considered as approaching 

 the same condition. Certain it is that close pollination occurs regularly in 

 this species. In cold weather, too, the flowers of N. amazonum seem to be 

 fertilized without ever rising to the water surface. The occurrence of 

 similar irregular cleistogamy in N. rudgeana may have been due to 

 unnatural conditions. Observations are totally lacking on Hydrocallis 

 species in their native haunts. 



In Castalia and Apocarpiae the flowers have a sweet ethereal odor. 

 This is richest and most delicious in N. odorata, faint in N. tetragona, 

 scarcely perceptible and only present on the first day of opening in N. alba. 

 The Lotos and Hydrocallis groups have very little odor ; in the latter it is 

 pleasant and ethereal, but in the former it is dull and rather offensive. 

 The flowers are always pistillate on the first day of opening, and the 

 stigmatic cup is filled nearly to overflowing with a large quantity of slightly 

 sweetened water. This is excreted by the stigma itself and by the axile 

 process. The stamens at this time stand erect around the stigma, leaving 

 a narrow vertical pit in the center of the flower, bounded all over the sides 

 by the still closed anthers of stamens of different lengths. All the stamens 

 are stiff in Lotos, but in Castalia the innermost filaments are very slender 

 and flexible, scarcely able to hold up the heavy anthers. The carpellary 

 styles also stand erect. In flowers of the second and later days the stigma 

 is quite dry and the papillae shriveled (in N. odorata and tuberosa ; 

 Robertson, 1889), though in N. alba, according to Schulz (1890) still recep- 

 tive. The carpellary styles, when long enough, have bent inward over the 

 stigma until they lie horizontally or point downward, and the inner 

 stamens meet in the middle so as more or less to hide the pistil. The 



