Annual herb, 10-30 cm tall, slender, sparsely branched from below, hairy with simple and branched hairs, sometimes subglabrous. Leaves distant, oblong or elliptic-oblong, sinuate-dentate to pinnatipartite, rarely subentire, 15-40 (-50) mm long, (5-) 10-15 mm broad, shortly stalked to subsessile; basal leaves laxly rosulate; upper smaller, oblong to oblong-linear, subsessile or sessile. Racemes lax, 10-15-flowered, increasing up to 10 (-15) cm long in fruit. Flowers c. 5 mm across, pinkish-violet to whitish; pedicels hardly 1 mm long in fruit, thickened. Sepals c. 2.5 (-2) mm long. Petals 6-8 mm long, c. 1 mm broad. Stamens c. 2:2.5 mm long; anthers c. 0.5 mm long. Siliquae 30-50 mm long, c. 1 mm broad, usually curved above, linear, subterete, subtorulose, sparsely pubescent to glab¬rous; stigma short but acute, bilobed; seeds c. 1 mm long.

Type: C. Asia: Balkash, Karelin & Kirilov 1240 (LE).

Distribution: C. Asia, Afghanistan, Iran and Pakistan.

Closely resembles Malcolmia scorpioides (Bunge) Boiss. but leaves different, flowers smaller, pedicel very short and fruits usually not so coiled.

Vasilcenko (in Kom., l.c. 282) has discussed the validity of the name Torularia brevipes (Kar. & Kir) O.E. Schulz (in Engl., I. c. 222) and invalidity of the name Malcolmia brevipes (Kar. & Kir.) Boiss. (l.c. 226), both based on Sisymbrium brevipes Kar. & Kir. (l.c. 155). For the latter he adopted the name Malcolmia karelinii Lipsky (1.c.), while earlier Schulz had accepted both, Torularia brevipes and Malcolmia brevipes as valid names but based on two different gatherings of Sisymbrium brevipes; the former on no. 106 and the latter on no. 1240. However, Boissier included both the gatherings under his Malcolmia brevipes based on Sisymbrium brevipes Kar. & Kir., and published it much earlier than Torularia brevipes O.E. Schulz (which perhaps, should not have been based on Sisymbrium brevipes Kar. & Kir., as it was already the basionym of Malcolmia brevipes Boiss., no matter partly or wholely) . Malcolmia brevipes (Kar. & Kir.) Boiss. is, therefore, a valid name, and if there is any dicrepancy it is with Torularia brevipes, published much later.