308 EXPEDITION TO JAPAN. 



gathered in the United States they would doubtless be referred to A. lyrata ; if in Europe, to 

 A. arenosa. 



Mattrtola annua, Siveet. Simoda. Cultivated. 



Cardamine impatibns, Linn. Simoda. A branching, pubescent form. Not mentioned by 

 Siebold and Zuccarini. 



Cardaminb macrophylia, Willd.; DC; Ledeb. Fl. Boss., l,p. 128. Simoda and Hakodadi. 

 The specimens are all corymbose at the summit, bearing from three to five racemes ; the flowers 

 apparently white, and the leaves are rather more downy beneath ; otherwise the plant exactly 

 accords with large-leaved forms of this striking species from Altai. The fruit not seen. This 

 also was unknown to Thunberg and to Siebold. 



Draba nemorosa, Linn. D. nemoralis, Fhrh.; DC; Sieb. & Zucc. I. c. Hakodadi and Yoko- 

 hama. Pretty large forms, but clearly the European plant. 



Capsella Bursa-pastoris, Moench. With the preceding. 



Brassica Chinensis, Linn. B. orientalis, Thunb. Low pine forests, Yokohama. 



Eaphanus sativus, Linn. Simoda. Mentioned by Thunberg as one of the commonest esculent 

 herbs of Japan. 



Polygala Japonica, Eoutt. ; DC Prodr. 1, p. 324. P. vulgaris, Thunb. Yokohama and 

 Simoda. A well-marked species, agreeing with the character given by DeCandolle, except that 

 the stems are assurgent. 



Viola palustris, Linn, f A single specimen from Simoda appears to belong to this species 

 (mentioned by Thunberg) ; but the leaves are cordate-ovate rather than reniform. 



Viola Japonica, Langsd. ex DC Prodr. 1, p. 295, not of Korthals, in Walp. Ann. 2, p. 66? 

 V. obovata, Thunb. 1 Yokohama. Specimen too incomplete to judge of. 



Viola G-meliniana, R. & S. ; var. glabra, Ledeb. Cat. Dorp., &FI. Boss. 1, p. 246? Simoda; 

 a single and insufficient specimen. 



Viola grypoceras (n. sp.): acaulescens, glabra ; foliis reniformibus et ovato-cordatis ; stipulis 

 lineari-subulatis longe setoso-pectinatis ; flore pallide cteruleo ; calcare adunco cylindrico 

 obtusissimo petalis omnibus imberbibus fere asquilongo ; stigmate nudo inappendiculato, rostro 

 brevi porrecto. — Yokohama, on wooded hill-sides ; March. Probably stoloniferous, but this is 

 uncertain. Leaves from half an inch to an inch in diameter, crenulate-toothed, thickly punctate 

 with brown dots, as in numerous species. Scapes filiform. Sepals acutish. Spur when fully 

 developed almost half an inch long, narrow, with a thicker and very obtuse curved extremity. 

 Anther-spurs very long, ligulate. This surely cannot be the V. Japonica of Langsdorff, nor, 

 with its long spurs, could Thunberg have mistaken it for V. odorata. The stipules resemble 

 those of V. adunca, Smith, (V. longipes, Nutt.) of Northwest America, although more densely 

 and strongly fringed ; but the plant is stemless, and the flowers, stigma, &c, very different. 



Viola canina, var. ? Japonica, DC, Prodr. 1, p. 298 ; Sieb. &Zucc. 1. c. Yokohama. Were 

 the specimens (having only a single blossom) sufficiently complete for determination, this would 

 doubtless be shown to be wholly distinct from Viola canina. The foliage is similar to that of 

 V. canina, but the stipules are much larger and strikingly fimbriate, in the manner of the pre- 

 ceding species, only more conspicuous. It is possibly a caulescent state of the last. 



Viola laciniosa, (n. sp.) : subpubescens ; caule valido folioso ; foliis ovato-cordatis obtuse acu- 

 minatis ; stipulis foliaceis magnis oblongis eximie pinnatifido-lacin:; tis snmmisve lanceolatis 

 superne integriusculis, laciniis lineari-lanceolatis ; pedunculis folium subsuperantibus ; petalis 

 ceeruleis, lateralibus hinc leviter barbatis, calcare brevissimo crasso scrotiformi ; stigmate dorso 



