<^g4 THE NATURALIST. 



i. SETIGERJ3. *' Turiones recti-aculeati simultaque setigeri." Many of tlie 

 species which he comprises in this section, are, it seems to me, far from 

 presenting this character. B. glutinosa, Sibth., with its pubescent glan- 

 dulose leaves is scarcely well placed in the same section which includes U. 

 lutea, Mil., R. spinosissima, L., R. alpina, L., R. sulphiirea, Ait., R. ferox, 

 Ait., and R. cinnamomea, L. ii. Aculeos.e. '* Turiones absque setis 

 aculeati." This section comprises five sub-divisions, established on the 

 form of the prickles and the clothing of the leaves : — Sub-div. 1. VilloscB. 

 Prickles almost straight, leaves villose : "Aculeis retiusculis, foliolis 

 moUibus." Several species in the first section ought to be placed here. 

 R. glandulosa, Bellardi, is wrongly put in this section since its leaves are 

 glabrous ! Sub-div. 2. Rubiglnosa. Prickles recurved, leaves glandulose 

 beneath : " Aculeis recurvatis, foliolis subtus etiam inter venas sparsim 

 glandulosis." A very natural section, but R. psilophylla, Rau., having 

 leaves without glands is strangely placed in the Rubiginosa. Sub-div. 3. 

 Canince. Prickles recurved, leaves glaiidless beneath, except a few on the 

 veins : " Aculeis recurvatis, foliolis subtus (costa quibusdam excepta) 

 eglandulosis." R. glandulosa, Bell., and R. psilopliylla, Eau., would be here 

 in their true place. Sub-div. 4. CentifolicB. " Aculeis difformibus, foliolis 

 regulosis." In the species it includes, this division is much confused, and 

 is not near so good as the two preceding sub-divisions in the connection 

 of its species. Here Reichenbach places R. marglnata, Walbr., R. Jundzil- 

 liana, Bess., and R. coriifolia, Fries. The two former having the prickles 

 uniform, and leaves with scattered glands beneath, have no affinity with 

 R. Gallica, L., and ought to be placed among the Rubigiuosce. R. coriifolia, 

 Fr., by its habit, prickles, and leaves belongs to the Canince, and not to the 

 section which includes Pi. Gallica. Sub-div. 5. NitidcF-. " Foliolis laevissi- 

 mis nitidis, stylis subcoherentibus quibusdam hologynis." R.fcetida, Bast, 

 which is found here, has not the styles united into a column, but free : by 

 its leaves glandulose beneath it is quite a stranger to any species of this 

 sub-division. Reichenbach if he had classed all the species of Rosa after 

 the sub-divisions of the section AculeoscB would have caused less interrup- 

 tion in the natural connection of the species. We may say, however, 

 this is one of the best classifications proposed up to now, spite of its imjDer- 

 fections. The prickles of the young stems is too variable a character to 

 be considered of primary value. 



Koch. — Synopsis Flora Germanicw et Helvetic^,'" (1843), employs 

 four sections for his genus Rosa, making use of the position of the carpels 



