102 II. G. SIMMONS. [sec. argt. exp. fram 



long, soft, lax hairs", and the fruit as "obovate, seeds not very numer- 

 ous, ovate, tumid, hairy, with short, awlshaped, reflexed beaks". Further 

 the leaves are said lo be "deeply cut, in a pedate manner into five, 

 seven or more, narrow, linear, obtuse, elegantly radiating segments", 

 which applies better to other allied species, than to that which I had 

 in my collection. I had occasion afterwards, also to see the original 

 specimens. There are in the herbarium of Linnaeus, four individuals 

 which doubtless belong to the same collection. On the paper is written 

 in the handwriting of Smith himself: "R. pedatifidus Sm. in Rees' 

 Cyclop, n. 72", and at one plant there is a sign *, which seems to be 

 older than the label of Smith and here is added "Siberia" (in pencil 

 by Smith)). 



Now the original specimens, as well as the description, show that 

 Smith has had in view, a plant very nearly allied to that which Brown 

 soon after described from Melville Island. There are, however, some 

 differences, that seem sufficient to separate the latter plant from that of 

 Smith. The characters that especially separate R. pedatifidus from R. 

 affinis are: the almost circular circumference of its basal leaves, that 

 are very deeply cut in a great number of very narrow, or almost entirely 

 linear, segments, with obtuse points, evidently grouped after a pedate design, 

 with a smaller middle, and two larger lateral, main lobes. The middle 

 lobe is only equipped with two lateral secondary lobes halfway from 

 the base, and these show a tendency to be again cleft. The lateral 

 lobes are deeply cut into at least three or four secondary lobes, that 

 only cohere a little above the basis, and are again deeply cleft. The 

 leaf is thus formed of a great number of narrow, or almost filiform, 

 segments. The basal leaves are more or less equipped with sparse 

 white hairs. The stem leaves are sessile or have a very short peduncle. 

 The lower ones resemble the basal leaves, but their segments are longer, 

 do not stand so close together, and become fewer in the upper leaves, 

 finally being only three. The sheath of the stem leaves half-embracing, 

 not so white-membraneous as in R. affinis. The plant is smaller and 

 more slender in all parts, the flowers are not inconsiderably smaller. 

 The white hair-covering of the stem is more prominent, and not reduced 

 only to the parts immediately below the insertions of the leaves. Especi- 

 ally the part of the stem below the lowest leaf is hairy. The flower- 

 stalk is, in the dried specimens at least, prominently canaliculate. 



The true R. pedatifidus, Sm. is, as far as I know, only found in 

 Asia, especially in Eastern Siberia and also in the islands of the Bering 

 Sea. A very imperfect specimen of Chamisso in the Kew herbarium, 



