and other new Surrey Plants, 85 



able to make any observations in that respect, but localities 

 have been noted in which each plant occurs alone, and I hope to 

 investigate the matter during the coming summer. The only 

 other character to which I need draw attention is afforded by 

 the female perianth scales — in ner/lectum these are narrowly 

 linear with a much broadened spathulate apex, while in ramosum 

 they are ligulate, thinner and more membranous, scarcely or 

 not at all enlarged at the apex. Those of the female flower are 

 excellently shown in the plate in Curt. Flo. Lond., but cm-iously 

 enough the male perianth scales of ramosum in the same plate 

 almost exactly resemble the female ones of neglectum. This I 

 hope to verify this year. The form of these scales is liable to 

 some variation, but the broad ligulate female scale is decidedly 

 characteristic of ramosum and the linear spathulate scale of 

 7iei)lectum. 



As S. neglectum has been repeatedly referred to S. simplex, — 

 that is, has been considered to have affinities with that species 

 rather than with ramosum, — I will conclude this portion of my 

 paper with a comparison of the two plants. On looking at a 

 head of fruit of each of these, a very considerable outward 

 resemblance is found in the pointed apex to the fi'uit and in its 

 long beak, but here the resemblance ceases. The colour of the 

 ripe fruit (a character which is of considerable value in this 

 genus) is quite different, as also is the shape. In S. neglectum 

 the broadest part is above the middle, and from this broad part 

 the fruit slopes off gradually by straight or slightly convex lines 

 to its base. In sbnplex the fruit is about equally broad at top 

 and bottom (oblong-fusiform, as Syme well calls it), with a slight 

 constriction in the middle ; the fruit is rounded below and then 

 contracted, so that the basal lines show a concavity. The endo- 

 carp of simplex is also much softer — rather woody than stony, — 

 being composed of larger, less dense cells, and is also without 

 prominent ridges. S, simplex has also not unfr-equently long, 

 semi-pellucid, floating leaves. A consideration of the characters 

 afforded by the leaves and fruit led me to see that the genus 

 falls naturally into two sections, and I afterwards found my 

 conclusions only corroborated those of the American botanists, 

 who follow a similar arrangement. The distinction made in 

 some continental works foimded on the sessile or stalked fr-uit 

 is of little value, this being a character liable to great variation. 

 I might say much more, in detail, in support of the American 

 system of dividing this genus into two sections, did time allow 

 of it. 



Since wi-iting this paper I have examined many plants in 

 flower, and the chief difference between them when in that state 

 appears to be found in the anthers, which in S. neglectum are 

 borne on much longer filaments than in .b>'. ramusum. The 



