'I"lu- Sihcriaii .s|u'ciincii.s liavc, on llic ollici hand, nunc (inadran^ular IcaNCS, willi 

 a siiuarc iiil hasc, only lait'ly and ([nilc (•\i(|)lionally somcwiial cordalely incised. 

 I'lie indcnialions ol Ilic lca\c.s arc li'wci in nnndx'i-. and I'oaiscr, j^cncrallv !! or 5. witii 

 miK'li deeper and jjioader ineisioii.s, and willi nearly cnneale or triangulai' leelli. 

 .siiliaeiile or subrolnndate al tlie apex, ^'enerally moie or less distinctly pointing up- 

 waids. 'i'lie shape of the leaves in both the s|)ecies is, for Die rest, lallier much varying, 

 and in Hie figure annexed I have brought together a fairly ritdi material from different 

 regions ol' Siberia and America in older to represent (he range of the variations of the 

 leaves. 



After observing this difference in the structure of the leaves, I have made a more 

 minute comparison between the Siberian and American s|)ecimens. 



There proved, indeed, to be a typical and ab.solutely constant difference in lin' 

 shape of the petals. These are very small, generally only from !'> lo ,') mm. long. Fig. 8!) 

 shows tile distinctions in the petals, and has been drawn from a great \aricly of speci- 

 mens taken in various localities. T'ig. a — i show petals of American specimens, fig. 

 n— t of various Siberian ones. It will appear that the size of the petals is much vary- 

 ing, and so is the proportion between length and breadth. There is, on the other iiand. 

 a marked difference in the shape of the petals in these two types. Thus, all of the 

 American specimens are seen to have all i)ut oliovate petals, broadest above the 

 middle, and broadly rounded at the lop, in Ihc lower pari gradually tapering, passing 

 by degrees into the equally broad claw. 



The petals of the Siberian si)ecimens, on the contrary, are pointedly ovate, sub- 

 acutish at the top, broadest below the middle, rounded at the base, and abruptly nar- 

 rowed into the claw. This shape of the petals is constant in all the Siberian specimens 

 I have examined. 



There is, according to the above mentioned, a difference so strongly marked, based 

 upon absolutely constant systematic characters, between the American specimens and 

 the Siberian ones that 1 have found it necessary to seperate them into 2 distinct species. 

 As PuRSCH has described the species from American material, the name of Ranunciiliis 

 Cymbalaria will have to be maintained for this one, the Siberian type being separated 

 as a distinct species: 



Raiiiiiiculus (Oxygraphis) siibsimilis nov. spec. 



Radix /ihrosa. Humilis. inntiuumoilo '2—6 cm. nlhi. glabra vd pilis sparsis 

 instructa, stolones repentes iiberius emitlcns. Folia ha.salia electa, longe petiolata. petioli 

 2 — 5 cm. longi. basin versus marginibus mrmbranaceis dilatati. Lamina .J— /.3 mm. 

 longa. glabra, in circuitu fere qiiadrala. basi plus minus rccle abscissa, superne leuiler 

 altenuata, aniice 3 — 5 denlibus grossis vel lobis instructa, imstice integerrima: lobi plus 

 minus distincte sublriangulares, basi lati.ssimi. superne cuneatini atlenu(di, apice 

 subobtusi vel subaculi. Caules floriferi erecli, teretrs, apbglli. foliis ba.filaribus aequi- 

 longi, rare foliorum longiludinem paulo superanles. vulgo simplires. unillori, rnro 

 ramosi, bi-vel irijlori. Diameter floris 6—9 mm. Sepala '/— .5 mm. longa. e.vcavata. 



239 



