142 THE ROSE FAMILY. [Agrimonb. 



whole of Europe, Russian Asia, and North America, but not an Arctio 

 plant. Frequent in England and Ireland, but becoming scarce beyond 

 the Clyde and Forth, in Scotland. Fl. all summer. It varies consider- 

 ably in the hairiness of the foliage, in the size of the flowers, and in 

 the form of the ripe calyx, which is more or less contracted at the base, 

 from obconical to campanulate ; and from this character two European 

 species have been distinguished, but the differences do not appear con- 

 stant enough to separate them even as marked varieties. [These are 



a. A. Eupatoria proper. Fruiting calyx obconic, deeply grooved, with 

 the lower spines spreading. 



b. A. odorata, Mill. More branched, with resinous odour, racemes 

 denser, flowers larger, fruiting calyx hardly furrowed, with the lower 

 spines spreading or deflexed.] 



XII. ROSA. ROSE. 



Erect, scrambling or climbing shrubs, more or less prickly, with 

 pinnate leaves, leafy stipules adhering to the leafstalk, and showy 

 flowers, either solitary or in small corymbs at the ends of the branches. 

 Calyx-tube globular or ovoid, contracted towards the top ; the limb 

 divided into 5 segments, often unequal, and sometimes lobed. Petals 

 5. Stamens numerous. Carpels several, 1-seeded, hairy, enclosed 

 within the tube of the calyx, which becomes succulent when ripe, and 

 sometimes slightly pulpy between the carpels, .the whole forming a 

 rather dry red or black berry. 



A well-marked genus, widely diffused over the northern hemisphere. 

 It comprises a considerable number of true species ; but several of them 

 being of very ancient and universal cultivation, and having been hybrid- 

 ised and multiplied with all the skill of modern horticulturists, their 

 more or less marked races or varieties are now reckoned by thousands. 

 Even in the wild state, endeavours have been made to characterise so 

 large a number of proposed species, that the confusion amongst them is 

 almost as great as in the Brambles. The forms indigenous to Britain 

 appear to be reducible to five types, which are probably real species. 

 It must, however, be admitted, that the characters separating them 

 are not so decided as could be wished, and that specimens will occa- 

 sionally be found that the most experienced botanist will be at a loss 

 to determine, and certainly not the less so if the number of British 

 species be extended to 11, as in Babington's Manual. 

 Prickles mostly straight, or very slightly curved, scarcely 



dilated at the base. 

 Stem seldom above a foot high when wild. Leafleta 7 or 9, 



usually small, and simply toothed 1. R. pimpinellifolin 



Stem 2 feet or more. Leafleta 5 or 7, usually doubly toothed, 



downy on both sides 2. R. villosa. 



Prickles, at least the larger ones, more or less curved, and 



dilated at the base. 

 Styles slightly protruding from the mouth of the calyx in a 



dense tuft, but not united. Stem scarcely trailing. 

 Calyx-tube globular, more or less prickly or bristly . . 2. R. vttlosa. 

 Calyx-tube ovoid or oblong, without prickles or bristles. 

 Leaflets very glandular, doubly toothed . . . . 3. JR. rubiginos*. 

 Leafleta without glands, or with very few on the edges 



only, simply or rarely doubly toothed . . . 4. R. canina. 

 ityles united in a column, protruding from the calyx. Stem 



very trailing ...... . 6. R. arvensit. 



