NOTICliS OJb* liOOKS AND MElMOIKS. 21 5 



Like H. Juranum, and H. (TothiciDii, it produces a rosette in seedlino- 

 plants, and in old x^lants, at the base of the flowering stems in late 

 autumn. In spring there is produced a more imperfect rosette, 

 and some of the leaves at the very base of the stem commonly 

 remain nearly to the time of flowering and sometimes even later, 

 especially in shady places. The seedling rosette, the late autumnal 

 rosette, the persistence of the lowest leaves, I have never met with 

 in wild or cultivated specimens of H. strictuni, H. crocatuni, 

 H. coryinbosuni, and H. prmumthoides. 



From H. Juranum it differs in having the stem scabrous, the 

 leaves much less amplexicaul and without distinct auricles, not 

 so conspicuously paler beneath and commonly more hairy above ; 

 the panicle when polycephalous does not produce numerous 

 branches termmating in small corymbs, the anthodes are con- 

 siderably larger, the perianth-segments not thickly clothed with 

 shoi-t, thick, gland-tipped hairs, and the ligules not distinctly 

 cihated at the apex. In the garden H. Dewari flowers about a 

 fortnight or three weeks later than H. Juranum. There is, how- 

 ever, a striking similarity between small specimens of the two in 

 which the branches of the panicle terminate in solitary heads ; 

 and I should not be surprised to hear that H. Dewari was the plant 

 Dickson had from Harehead Wood, Selkh'k, and that he sent 

 H. Juranum to Mr. Borrer from his own garden, believing it to be 

 the same as the wild plant. 



H. Dewari bears some resemblance to the broader-leaved states 

 of H. gothicwn, but differs in the leaves being semi-amplexicaul, 

 with a greater tendency to be ovate in outline and less denticulate 

 on the margms. I have never seen it with the distinctly- toothed 

 leaves so frequent in H. f/othicwn. The leaves are of a much 

 brighter green, and not so much paler beneath; they are also more 

 hairy. The j)anicle when polycephalous has not the same tendency 

 to produce branches terminating in corymbs. The anthodes are 

 considerably smaller, the inner phyllaries less attenuated and much 

 more liau\y. 



From H. strictum it differs (in addition to the points already 

 stated) by having the stem scabrous, the leaves broader in the 

 middle and more attenuated at each end ; and more conspicuously 

 by the long straggling branches of the polycephalous panicle. 

 But the greatest stress must be laid on the fact of its producing a 

 true rosette. All the specimens Mr. Backhouse has gone over in 

 herbaria I have access to he has named H. strictum. 



H. cory)nhosum, Fries, and H. crocatum, Fries, differ still more 

 widely from H. Dewari than does H. strictum. Both, especially 

 H. cori/mhosum, have the stem leaves much more numerous, more 

 parallel- sided, and the polycephalous jDanicle with the branches 

 ending in small corymbs. H. crocatum also has the pericline 

 much larger, more abru^Dt at the base, and with far fewer hairs. 

 H. cori/mbosum has the leaves more glabrous, and neither the one 

 nor the other produces rosettes. 



At one time I thought it might be //. datum, of Fries; but as I 

 now possess Lindeberg's 'Hieracia Scandinavian Exsiccata,' I can 



