122 THE WATERLILIES. 



stamens now dehisce. Where the flower is open only for two days, as 

 in Xanthantha, all of the stamens dehisce on the second day, but the outer 

 ones precede the inner by a few hours. In the rest of Castalia the inner 

 stamens dehisce on the second day of opening ; they do not completely 

 cover the entrance to the stigma from above, but they hang over far 

 enough to drop their pollen upon it, and sometimes close fertilization 

 occurs. In Brachyceras, as a rule, the entrance to the pistil is totally 

 obstructed on the second day by the inner stamens, and the outer ones are 

 dehiscing ; subsequently the inner stamens bend outward and dehisce 

 until, on the last day, the passage to the stigma is wide open again. In 

 spite of this, N. caerulea and stellata are said by Caspar)' (1877) to be 

 capable of self-pollination, the latter while still in the bud. The stiff 

 stamens of the Lotos group dehisce simultaneously on the second day, 

 and undergo no considerable change of position. 



The exact manner of effecting cross-pollination and the insects 

 concerned are unknown for the Lotos group, but the white and light-pink 

 varieties are extremely fertile when grown out of doors in the vicinity of 

 Philadelphia. The agency of insects in Hydrocallis is likewise undetermined. 

 N. alba is, in the opinion of Delpino, fertilized by scarabaeid beetles (Cetoniae 

 and Glaphyridae); he regards Nymphaea and Victoria as especially adapted 

 for beetles. Kirchner, in his "Flora von Stuttgart," page 275, mentions 

 "flower-beetles" as visitors {Jide M tiller, 1883, p. 93; Schulz, 1890), and 

 Picciola found Donacia (a chrysomelid beetle) abundantly. But Schulz 

 found N. alba perfectly fertile with its own pollen, and considered that the 

 flies and beetles which visit it always cause close fertilization. N. tuberosa 

 is visited by small mining bees (Andrenidae), flower-flies, and beetles 

 (Robertson, 1889), and N. odorata by Halictus and others. In our own 

 experience, mining bees and flower-flies are the only insects found. They 

 are drowned in large numbers in the stigmatic fluid, and are found in the 

 closed flowers after anthesis. In a single bloom of TV. capensis X zanzi- 

 bariensis I once counted thirty-two little bees ! Bacon (1874) noticed dead 

 insects in N. odorata and supposed that they were caught by the closing of 

 the flower ; Delpino attributed their death to the heavy odor of the flower, 

 and Planchon to the accumulation of carbon dioxide in the floral cup {fide 

 Robertson). But Robertson is undoubtedly correct in considering it 

 simply a case of drowning. The course of events is about thus : The 

 insects which visit the flowers are in search of the copious pollen ; no true 

 nectar is secreted. The earliest flowers to open each day are the older 

 ones, in which pollen is plentiful and the stigma is dry. The insects fly 



