37'2 ICOSANDIIIA— POLYGYNIA. Rosa. 



compressed, is perhaps of the next degree of importance. 

 I do not so much regard the frequent inequality of these 

 prickles. The transition of the upper stipulas into brac- 

 teas is nearly general in British Roses, and therefore less 

 discriminative than Mr. Woods in some cases makes it. 

 The simple or double serratures of the leaflets, and even 

 their pubescence, are in my opinion duly appreciated by 

 this judicious writer. The characters derived from 

 hence do indeed occasionally vary, as is the case, more 

 or less, with every botanical distinction. There is no 

 mathematical certainty in natural history, our definitions 

 being generally but a choice of difficulties. The occa- 

 sional failure in particular instances, even of specific 

 characters generally most decisive, as opposite, or alter- 

 nate, entire or serrated, leaves, does not overset our ge- 

 neral confidence in those distinctions. So neither does 

 a partial or occasional change in the leaves of Roses, 

 from simple to compound serratures, prove such differ- 

 ences entirely futile. I am led therefore, by a careful 

 revision of the genus, to think the accidental mutations 

 to which I allude have induced Mr. Lindley to combine 

 too many species together ; while Mr. Woods, by too 

 absolute dependance on some characters, has perhaps 

 now and then made species of varieties. In the study 

 of a genus hitherto imperfectly understood, the latter is 

 the least injurious error. Corrected judgment may, 

 hereafter, combine what precise observation, in the first 

 instance, has with due caution separated, and my spe- 

 cific definitions may then be curtailed. I trust that 

 neither of my intelligent friends will feel offended if, in 

 a matter of so much conjecture, uncertainty, and no- 

 velty, I, in a few instances, differ from them. What we 

 have done may afford a clue for the guidance of others, 

 who will try all our principles, by applying them to 

 practical use. In the synonyms of foreign authors, un- 

 less verified by a comparison of specimens, I have been 

 very sparing, seeing how many mistakes are daily made 

 by compiling or copying, instead of observing. 



* Branches bristlij. Prickles mostly slender, nearly straight. 



t 1 . R. cinnamomea. Cinnamon Rose. 



Flower-stalks bracteated, smooth. Calyx mostly simple. 

 Fruit globose. Bristles deciduous. Leaflets oblong, 

 simply serrated, finely downy; glaucous beneath. 



