13G THE JOUKNAL OF INDIAN BOTANY.
than those of dry ground ; and it is therefore not surprising that there is a sameness about the vegetative parts which we do not find in dry land plants. Thus in Eriocaulon the stem is nearly always very short, little more than a flat disc, with the leaves all radical and narrow, and the flower-heads are carried well above the ground or the water- level on tall scapes. There are of course differences ; submerged species have linear leaves, and some have elongate stems ; some land species are hairy, most glabrous ; in some species there is only one, in most there are numerous scapes : but except for these the differences are mostly small and difficult to define. This sameness in the vegeta- tive parts is accompanied by a surprising amount of difference in the floral, on which therefore the separation of the species is of necessity largely based. But since the parts are always very small and need a good lens for their study the species are difficult to diagnose, and in most cases it is impossible to do so with the naked eye without con- siderable practice. Collecting therefore, for any one not thoroughly conversant with the species, is like pulling things out of a bran-tub, one cannot tell at once what one has got ; and the determination of the species afterwards, unless the descriptions are very clear and to the point, is difficult in the extreme. This is borne out by an exami- nation of the material in the Indian Herbaria. Thus six collections made by Meebold in Mysore and Coorg, from September 1897 to November 1898, of what is certainly one species, were named by him, E. sexangulare three times, E. truncation twice and E. trilobum once. Yet these three species should never be confused, and the plants col- lected did not as a matter of fact belong to any one of them. The name E. sexangulare I find wrongly given on 15 collectings in the Herbarium 'of the "Royal Botanic Gardens, Calcutta, belonging to seven species quite distinct from the true E. sexangulare of Linnaeus ; and E. luzit- laefolium to 13 collectings belonging to six species, four of them being among the seven just referred to ; and these are not exceptional cases ; the third column of the Appendix will show that 2 species have each been given names of 7 other species, 2 more names of 6 other species, 7 of 3 or more other species.
One result of this is that species have bean given a wider distri- bution than they are entitled to, and this has of course led many a collector to suppose that he may have a species which really does not exist in his area. Especially is this the case with E. luzulaefolium Linn, which has been reported as all over South India, from Khasia to Bombay, Madras, Malabar and Ceylon. But, unless the sheets which bear Wallich's number in the Calcutta Herbarium do so wrongly, the species is confined to Nepal, Assam, Bengal and Upper Burma.
