Pkakger — On Species of Sediim collected in China. 3 



iniiiutae, cuneato-triuicatae, '6 mm. longae. CarpeUa 4 mm. longa, erecta, 

 rubi'opunctata, in stylos erectos attemiata. 



Kiosliaii and vicinily, on l)ills, province of Honan. I.at. about 33'. June 23, 

 1917. (L. 11. Bailey, Sedum no. 5.) 



This is a very interesting plant, both on account of its peculiar foliage 

 and its relationships. Its fantastic leaves are without parallel in tl:e genus 

 to wliich it belongs. The white liorny spined sheaths which envelop the leaf- 

 tips are very persistent, and clinging to the old rotting leaves, are strikingly 

 suggestive of small white Crustacea or arachnids crawling on the plant. 



In its dense rosette of well-developed leaves producing lateral leafy 

 flowering stems the plant is very exceptional among Sedunis. But a similar 

 arrangement is found in S. Balfouri R. Hamet from the Yangtse-Mekong 

 divide, in <S. Durisi E. Hamet from Zumutch 'J agh in Central Asia, and 

 also in <S'. orichalcum W. W. Sm. from mountains north-east of the 

 Yangtse bend. 



I have endeavoured elsewhere^ to show that a continuous series of forms 

 connects the typical members of the Bhodiola section of Sedum (such as 

 roseum Scop., cvassipes Wall., himalense I). Don), which possess an elongated 

 caude.x crowned with scales from the axils of wliicli leafy Hower-shoots arise, 

 with such forms as Balfouri, in which a dense rosette of linear caudex-leaves 

 produces tall axillary flower-shoots. The scales which crown the caudex of 

 the one are analogous to the leaves which crown the caudex of the other, all 

 intermediate stages being observable among the different species, and seedlings 

 of the scale-bearing species producing leaves (which early degenerate into 

 scales) exactly analogous to the leaves of the leaf-bearing species. One of the 

 strongest links in the continuity of this series lies in the mode of attachment 

 of the leaves. In most Sedums, although the base of the leaf or leaf-slalk 

 may be (and often is) broad, the actual attachment is very contracted, and 

 little more than a point. In the Ehodiolas, however, while the attachment 

 of the leaves of the flower-shoots is as just described, that of the caudex- 

 leaves is very broad, the leaves clasping the caudex and being attached by 

 their whole breadth. Precisely the same arrangement occurs inS. limuloidcs 

 (PL I, h, c), and its leaf also agrees generally (except for its peculiar apex) 

 with that of S. Balfouri (Plate I, h). 



On the other hand, S. limuloides produces, in addition to its axillary flower- 

 stems, a terminal flower-stem, following the maturation of which the whole 

 shoot, incliiding the basal leaf-rosette, appears to die, the life of the plant being 



' On the affiuitios of Sedum Fraeyeria)iutn W. W. Siii., with a tentative clasBitication 

 of the section Mhodiula. Pioe. Bot. Sue. Ediub. 27, pi- -, I'JIT. 



