146 THE JOURNAL OF INDIAN BOTANY. 



Mountains from Mt. Aboo to Coorg, but do not go further north, 

 east or south. 



The group collinum-quinquangulare-trilobum-Dianae, has collec- 

 tively the widest distribution, and shows very interesting develop- 

 ments and cross-relationships. E. quinguangulare belongs to the 

 plains of Ceylon, S. India and the Deccan, extending only rarely to 

 the Western side in Canara. On the mountains to the south it is re- 

 placed by E. collinum Hook. f. and to the north in Bengal by E. tri- 

 lobum Ham,, both of which hardly differs from it except in the much 

 darker head and better developed sepals, but are really quite easily 

 distinguished. E. quinguangulare and E. trilobum have their counter- 

 part on the wastern side in E. Dianae Sp. Nov. which in its widely 

 differing varieties is similar to both, but differs in the reduction of one 

 sepal to generally only a bristle and in the involucral bracts being 

 usually longer. E. collinum has likewise in Ceylon one sepal smaller 

 than the others, but the forms are not otherwise distinguishable. 

 Both E. quinquangulare and E. Dianae show lengthening of the invo- 

 ucral bracts, though the former only in Burma ; and E. xeranthemum 

 with sometimes 3 female sepals sometimes 2, might be derived 

 from either. It is in fact as if the two species E. quinquangulare 

 and E. trilobum, and perhaps also E. collinum, were originally one, 

 and developed from it as varieties and later on species, in — (l) the 

 plains of S. India : and L. Bengal, (2) Upper Bengal, and (3) the 

 mountains of S. India : and further as if these have independantly 

 suffered a reduction in one sepal of the female flower ; the first two 

 in travelling westwards across the Ghats to the sea, the third in 

 crossing over to Ceylon. Another change was a lengthening of the 

 involucral bracts, which seems to have proceeded independently in 

 both E. quinquangulare and the derivative E. Dianae, as it has done 

 also in other species. 



The Mendelian would doubtless find in crossing a sufficient ex- 

 planation of these double relationships, but it remains that the 

 species as here defined occupy distinct areas and are found together 

 if at all only on the borders of contiguous fields. 



Reference to published Works. 



F.B.I.— Flora of British India, by J. D. Hooker. Volume ; VI, (1894) 



pages quoted in Arabic numerals : a species number given 



thus, No. 3. 

 Ruhl, — Die Eriocaulaceae by W. Ruhland in Engler's Das Pflanzen- 



reich (1903). The serial number of the species alone is 



quoted. 



