Taxonomy Lotos. 



193 



Notes. The following species differ chiefly in color of leaf and flower, in shape 

 of leaf, and in production of seed. It is stated in Hooker & Thomson (1855) that 

 all gradations in color occur in India from the white to the deep-red forms; they 

 therefore united them into one species, N. lotus L., with three varieties : a. Lotus; leaves 

 orbicular or reniform, densely pubescent beneath ; lobes divergent or approximate ; 

 flowers ample, red, rosy or white. /9. cordifolia; leaves cordate-ovate, densely pubes- 

 cent beneath ; lobes divergent ; flowers of medium size, white or carneous. y. pubes- 

 cens; leave, puberulent or pubescent beneath ; flowers smaller, white, rosy or red. 

 These characters are wholly unreliable in cultivation, whereas the color relations come 

 pretty true to seed, and are usually reproduced precisely from the tuber. Caspary's 

 (1865) division of the species into var. lata with sepals ovate to ovate-oblong, petals 

 white or rosy, filaments yellow; var. angusta, with sepals mostly oblong-lanceolate, 

 petals white, inner filaments with a purple spot; var. oblonga, with sepals mostly 

 oblong, rarely oblong-lanceo- 

 late, petals white, rosy or pur- 

 ple, is also unsatisfactory. 

 There seems to be no con- 

 stant morphological differ- 

 ence on which to base a clas- 

 sification. In 1888, however, 

 Caspary speaks of N. lotus, 

 pubescens, and rubra as dis- 

 tinct species. This opinion 

 may have been founded on 

 the relations of hybrids, of 

 which many are in cultivation. 

 But the evidence from this 

 source is not conclusive. For 

 example, the hybrid N. oma- 

 rana has bronzy green leaves 

 and large flowers with petals 

 pinkish-red with a nearly white streak up the middle, and the floral leaves open 

 out horizontally ; it shows plainly the blending of its N. lotus and N. rubra parentage ; 

 it bears seed, however, much more freely than N. rubra, but less freely than N. lotus. 

 Its exact parentage is stated to be N. lotus X N. sturtevantii; the latter is a seedling 

 from N. devoniensis, raised without crossing, according to Mr. Sturtevant, but it 

 gives evidence of N. lotus parentage. N. devoniensis was at first considered to be a 

 hybrid of N. lotus and N. rubra, but it shows no evidence of white parentage, and 

 is probably not of hybrid origin, and Hooker & Thomson claim to have found it 

 common in India ; it very rarely produces seed, however, a character which it seems 

 to have inherited in exaggerated form from its red parent. Whatever else may be 

 inferred from these facts, certain it is that crossing of these forms does not induce 

 sterility, but rather that fertility increases the nearer the mixed progeny approaches to 

 the highly fertile N. lotus. From what we have seen, however, there seems sufficient 

 reason to consider N. lotus, pubescens, and rubra quite distinct. 

 15 



Fig. 73. Floral organs of the Lotos group. Outer, median and inner 

 stamens; carpellary style; vertical section of ovary. Natural size. 

 From a white flowered garden form. 



