COMPOSIT-^. 19 



Var. 3 is a remarkable plant, Avliich probably ought to be con- 

 sidered a sub-species, but which is retained as a variety because 

 tliore appear to be on the Continent several intermediate forms 

 (which I have not seen) wliich seem to render it impossible to draw 

 any line of demarcation between it and the common form. In this 

 form the leaves are not only flat and less deeply divided, but the 

 spines are so weak that the leaves may be handled with impunity, 

 and they are all much more attenuated at the base. The anthodes 

 too, in the specimens which have come under my notice, are 

 arranged in a long lax panicle. 



Creeping Pliune-Thistle. 



French, Herbe-aux-varices. German, Feld Kratzdistd. 



nYBPtlDS. 



CARDUUS PRATENSI-PALUSTRIS. 



Plate DCXCV. 

 Onicus Forsteii, Sm. Eng. Fl. Vol. III. p. 390. 



Intermediate between C. pratensis and C. palustris ; diCfering 

 from the former in the leaves being less decurrent, so that the 

 stem is not so continuously winged ; the under side of the leaves 

 generally slightly arachnoid-pubescent ; the stem, especially the 

 upper part, cottony ; the anthodes larger and more globose, less 

 aggregated, usually not more than 2 or 3 together ; the rootstock 

 perennial ; the phyllaries with shorter and weaker spines. From 

 C. pratensis it is distinguished by the radical fibres being more 

 slender ; the flowering-stems growing more in tufts ; the leaves 

 more deeply sinuate-pinnatifid and spinous, those in the middle 

 of the stem slightly decurrent, not hoary-white beneath ; the an- 

 thodes much smaller and less arachnoid, the outer phyllaries with, 

 an evident spine. 



In boggy places. Very rare. Frant, two miles from Tun- 

 bridge Wells (]\Ir. S. F. Forster), and White Moor Pond, near 

 Guildford, Surrey (Mr. H. C. Watson) ; near Fridge and on Ditch- 

 ling Common, Sussex (Mr. Borrer). 



CARDUUS ACAULI-PRATENSIS(?). 



Plate DCXCVI. 

 C. Woodwardii, //. C. Watson, Cyb. Brit. VoL II. p. 83. 



Intermediate between C. acaulis and C. pratensis, and varying 

 extremely in appearance, but always more nearly resembling the 



