1 74 THE WATERLILIES. 



No. 1283. Franzensbrunn : 3 Nomenteiche, coll. Caspary, Aug. 25, 1855, in hb. Kew. 

 Lakes of the Vosges near Sembach, Bavaria, coll. Koch, July 20, 1859. Hammertcich 

 near Falkenberg, Ober-Schlesien, fid. specimen in hb. Kew. Teplitz in Bohemia, coll. 

 Winkler, 1853, in hbb. Boissier, DeCandolle. Near Stockholm, Sweden, coll. Ljo- 

 gruen, 1845, m hbb. Munich, DeCandolle. Lithuania, coll. E. Rudominowna, in hb. 

 Kew. Neusattel near Ellnbogen, coll. Caspary & Ortman, 1855, in hb. Kew. Schwarz- 

 see near Kitzbiihel, Tyrol, coll. Traunsteiner, No. 1608, marked " N. alba," in hbb. 

 Boissier, Paris. Lakes of the Baikal region, Siberia, coll. Fischer, 1836, in hb. Paris. 

 Cashmere Lake, coll. J. E. Winterbottom, May 13, 1847, and coll. Falconer, both 

 marked " N. alba" in hb. Kew. 



NOTES. The species varies widely as indicated above in the description, and as 

 may be inferred from the lengthy synonymy. Presl's distinctions were not widely 

 known, or were lost sight of by most botanists. He separated N. Candida from N. alba 

 largely on account of the absence of stamens from the summit of the ovary. The 

 same name had been used for a European white waterlily by Tragus, Fuchs, Caesal- 

 pinus and others. N. paitciradiata was characterized by its small flowers and few car- 

 pels (8 to 9). N. biradiata was noticed on account of the blood-red star on the 

 stigma. N. bastiiniana was separated from N. alba on account of having oval leaves 

 instead of round, and its small number (9) of carpels. N. cachemiriana has also only 

 7 to 8 carpels ; the original description is quite complete. N. scmiaperta has as its out- 

 standing trait the fact that even at noonday the flowers remain only half open. N. 

 ncglecta was marked off from alba by its small oval fruit, few rayed stigma, and the 

 converging course of the lowest pair of veins of the leaf; from Candida in having 

 petals longer than the calyx, and ovary only one-third bare, instead of two-thirds as 

 in Presl's Candida; from scmiaperta by smallness of ovary and color of stigma. It is 

 interesting to notice how various observers have been impressed with different feat- 

 ures of the plant as they appear more or less exaggerated in certain localities. 



Probably the best disposition of the case is that of Caspary in 1879, when he first 

 granted specific recognition to this group of forms. His paper of 1855 placed these 

 plants in variety oocarpa of N. alba, with the following divisions : 



N. alba var. oocarpa. 



E. angustata styles longer and narrower; Konigswerth in Bohemia. 



F. scmiaperta styles shorter ; flower opening one half. 



a. crythrocarpa fruit reddish (=N. scmiaperta and biradiata) 



b. chlorocarpa fruit greenish (= N. ncglecta) 



G. aperta: 



a. erythrocarpa (=N. kosteletzkyi and cubogermen, from Franzensbrunn) 



(= N. Candida Presl from Pardubitz in Bohemia) 



b. chlorocarpa (=N. Candida Ortm. from Ellnbogen, Bohemia) 



(= N. Candida Presl from Pardubitz, Bohemia) 



The name oocarpa was changed by the author to oligostigma in 1863, and finally, 

 on discovering that crosses between N. alba oocarpa and N. alba sphacrocarpa were 

 true hybrids, with imperfect sexual apparatus, the two were made distinct species, and 

 the following disposition of the one in hand was given, with a few, mostly Swedish, 



localities. 



