146 THE WATERLILIES. 



the close resemblance of N. caernlea and N. stcllata, and partly in the fact that these 

 and N. capensis were introduced into cultivation near the same time, and the descriptive 

 character of the name caerulea made it equally applicable to all. The geographical prox- 

 imity of this and TV. capensis may have added to the confusion. Indeed, examination 

 of herbarium material shows that N. caerulea is connected by intermediate forms with 

 N. micrantha on the west and N. capensis on the south. We have not been able to 

 secure living material of N. micrantha, but the capensis type is very easily distin- 

 guished from the present species. Their hybrid (N. pulcherrima Tricker) is entirely 

 sterile. Other related forms are discussed in their proper places. Many botanists have 

 included all the blue waterlilies of Africa and Asia under the name of N. stellata, as 

 did Hooker & Thorns. (1855) ; Caspary, ten years later, separated off N. capensis, and 

 in 1877 gave exact descriptions of four species in parallel columns; he stated that his 

 old arrangement (1865) needed complete revision, but unfortunately this work was left 

 undone. 



The Kotschy collection from Nubia has long been a stumbling block to students 

 of Nymphaea. They are marked No. 166 N. caerulea Sav. and No. 167 N. anipla 

 DC. " Kotschyi Iter Nubicum. In stagnibus pluvialibus ad radices eurontas montis 

 Cordofani, Arasch Cool, Jan. 5, 1841, d. Oct. 12, 1839." The first is represented in hbb. 

 Kew, Paris, Boiss., Deless., and Berlin. It was marked by Caspary "N. capensis 

 Thunb." The Boissier specimen has narrower petals and fewer stamens than the 

 type of capensis; the Berlin specimens differ between themselves one flower has 

 spotted sepals, and the leaf gently sinuate ; another has leaves typical of N, caerulea 

 but no spots on the sepal. No. 167 was marked "A", caerulea Sav." by Caspary in hb. 

 Boiss., with the addition of " var. albida " in hb. Paris. The leaf is gently sinuate in 

 the Kew, Deiessert, and one Paris specimen. It is also represented in hbb. Berlin and 

 Boissier. The leaf may or may not have black spots and some purple coloration beneath. 

 The flower resembles that of N. caerulea, but was evidently white, and in the Kew 

 specimen, at least, the outer stamen is provided with an unusually long, lanceolate ap- 

 pendage. The two plants seem strikingly intermediate between N. capensis and N. 

 caerulea. No. 167 is the type of Lehmann's N. nubica. 



Self-fertilization occurs regularly in this species. 



Reference to the importance of N. caerulea as the Blue Lotus, the lotus par excel- 

 lence of Egypt, has been made in Chapter I. 



Nymphaea micrantha Guillemin & Perrottet. (Plate IX.) 



Leaves small, round-cordate, peltate, entire, petiolate ; petioles very long, slender ; 

 lobes divaricate, acuminate ; under surface reddish, marked with violet-black dots ; upper 

 surface glabrous, pale green, in mature plants bulbiferous at the summit of the petiole ; 

 stigmatic rays 15. 



N. micrantha, Guillemin & Perrottet 1830. Hooker 1850. Paxton 18530. 



N. caerulea, Guillemin & Perrottet 1830. Not Savigny 1802. 



N. caerulea var. albida, Guillemin & Perrottet 1830 = A : . micrantha, Hook. 



N. rufescens, Guillemin & Perrottet 1830. 



N. stellata var. bulbillifera, Planchon 1852 c. Not N. stellata, Willdenow 1797. 



N. vivifara, Lehmann 1852 a, b, 1853 a, 1854. 



N. gitineensis, Schumacher & Thonning 1829, fid. Chifflot 1902. 



