Mr. Winch's Observations on his Flora. 141 



begins to show a tendency to pubescence on the midrib of its leaves, and 

 Rosa dumetorum or cassia has leaves covered with a velvety down. In 

 all these the prickles are strongly hooked. The next link of the chain 

 claims particular attention, as it appears to connect two sections of the 

 genus. I allude to Rosa tomentosa, the prickles of which are in part 

 hooked, and in part straight, its fruit sometimes smooth, at others 

 hispid, and this frequently occurs on the same bush, but its leaves are 

 always pubescent. Rosa scabriuscula has straight prickles, its flowers 

 are white, except the extremity of the petals which are tipped with red. 

 Rosa villosa, of Smith (not the plant so called by Swartz and most of 

 the continental Botanists, that being an exotic and the Apple Rose of the 

 gardeners), has invariably straight prickles, pubescent leaves, globose 

 fruit, occasionally hispid, and full rose-red flowers. The Roses most 

 nearly allied to Rosa canina have pale pink blossoms, those to Rosa vil- 

 losa red, with the exception of Rosa scabriuscula as previously mention- 

 ed. With respect to Rosa Eglanteria, it will be found to vary through 

 Rosa micrantha to Rosa Borreri, or inodora of the foreign Botanists, a 

 bad name for a plant which still possesses, in some degree, the odour of 

 the Eglantine, and which the old herbalists would have denominated a 

 savage Sweet Briar. Passing on to another section of the genus, we 

 find the same uncertainty of specific character to continue. Thus in 

 Rosa gracilis of Woods, the prickles on the stem are hooked, but not so 

 on the young branches, — in its variety Rosa Sabini or Doniana (for there 

 is not the slightest difference between them), all the prickles are straight, 

 as well as in its dwarf sea-side variety, Rosa involuta. Rosa spinosis- 

 sima, in its wild state, varies only in size, though the parent of the nu- 

 merous beautiful Roses sold by nurserymen under the name of Scotch 

 Roses, and Rosa arvensis continues true to its well-defined specific cha- 

 racter. Of Rosa rubella, which is nearly allied to Rosa alpina of the 

 south of Europe, as Rosa arvensis is to Rosa sempervirens — too little is 

 known to enable one to say whether it passes into any other supposed 

 species — its pendent scarlet fruit keeps it asunder from any British 

 Rose. A dark red flowered variety is not rare in the gardens near 

 London, but I could not learn from whence it came. The foregoing 

 vol. n. o o 



