Perennial herbs, often tufted and pilose with simple or forked hairs, very rarely glabrous or subglabrous, often suberect or procumbent, branched mostly from below. Leaves spathulate or obovate-oblong, (1-) 3-5 (-10)-toothed or lobulate above, ± entire below, sessile or subsessile; basal leaves rosulate, somewhat larger than the stem or upper leaves. Racemes few to many flowered, often bracteate below, sometimes scapose. Flowers white or lilac, rarely mauve or yellowish with a lilac tinge inside; pedicels often turning towards one side, pilose. Sepals subequal to unequal, free or connate below, not or hardly saccate at base, sometimes persistent. Petals about twice as long as the sepals, spathulate. Stamens 6; filaments not appendaged. Lateral nectar glands annular; middle glands joining the laterals. Ovary linear to elliptic-ovate, often pilose, few to many ovuled; stigma usually depressed, subretuse, sessile or subsessile (very rarely on a short style). Fruits (usually siliquae) ovate-oblong, lanceolate, linear-oblong to linear, flattenea, usually bilocular and tardily dehiscent, pilose to glabrous; septum membranous, thinnest in the middle, complete, incomplete or obsolete; seeds 1-2-seriate, few to many, mucilaginous or not when wet; radicle accumbent rarely incumbent.
Nearly 20 species, chiefly in C. Asia, Tibet and Himalayas; 7 species are recorded from our area.
A controversial genus sometimes split into a number of genera, but habit, leaf, flower and fruit characters are fairly constant. Presence of accumbent and incumbent radicle in one tribe or even in one genus, is not a so rigid rule (see Arabis and Phaeonychium in this tribe) and to separate Christolea and Ermania just on the basis of incumbent and accumbent radicles respectively, is not taxonomically sound and that too under different tribes. Soviet taxonomists, no doubt, have better access to the material of most of the species of this genus, but it appears that they are hardly or not justified in recognizing genera like Eurycarpus Botsch., Ermania Cham. ex Schulz. (s. str.), Desideria Pamp., and Oreoblastus Suslova based on species drawn from this genus. Status of Parryopsis Botsch. and Vvedenskyella Botsch. or at least some of their species, needs revaluation.
However, this seems to be one of the difficult spots in this family taxonomically where generic, and to some extent tribal boundaries seem to be quite superfluous. Ermania now seems to have left with only one species, Ermania parryoides (Cham.) Cham. ex O.E. Schulz (Draba parryoides Cham.) with n =6 and perhaps others have n=7 (Suslova in Bot. Journ. U.R.S.S.57; 647-653. 1972; Botsch. in Bot. Journ. U.R.S.S. 57: 668.1972). After much thought, with this complex, I have come to the conclusion that it is difficult to maintain the separate identity of Ermania, Desideria, Eurycarpus and Oreoblastus from Christolea, where most of our area species are involved. Desideria is said to have connate sepals and obsolete septum in linear fruits and Eurycarpus has septum absent or obsolete in ovate fruits but these characters can hardly be called as generic characters, looking into the entire range of variations of the taxa involved. Even if, Christolea Camb. (s. str.) and Ermania Cham. ex Schulz. (s. str.) are considered as separate genera, there can be no doubt that Oreoblastus Suslova is congeneric with Desideria Pamp., where most of our species would go.
Furthermore, the genus Christolea cannot be retained under the tribe Sisymbrieae with the type of flower and fruit it has. Rechinger f. (l.c.) also inadvertantly, placed his genus Koelzia (which is congeneric with Christolea sens. str.) under the tribe Arabideae. However, Suslova (l.c.) retains Ermania (s.str.), with a single species, in the tribe Arabideae, while includes most of the “species” of Erimania (s. lat.) in his new genus Oreoblastus and places it under the tribe Matthioleae, leaving Christolea (s. str.) under Sisymbrieae. This camplex needs further investigations.