174 BULLETIN DE l' HERBIER BOISSIER. (13) 



This seems a very distinct plant; it combines Ihe foliage characteristic 

 of C. Ledebouriana with flowers very like Ihose of C. macrocentra and 

 C. Boissieri. In the St. Petersburg collection Ihere are however a nuniber 

 of spécimens that appear referable to C. Ledebouriana which have the 

 spur, though a good deal shorter, upturned as in C. cyrtocentra. The 

 spécimens on which the species is founded are preserved in the her- 

 baria of Saharanpur and Calcutta. 



12. Corydalis persica Cham, et Schlecht. Linnsea 1, 567 (1826); 

 foliis caulinis sessilibus forsan ad normam 3-sectis, in specimine unico adhuc 

 receplo tarnen folium alterum segmente perfecto mediane cum segmente 

 laterali minore singulo, alterum segmentis lateralibus subsequalibus 

 mediane déficiente gaudere videtur, segmentis omnibus longe petiolu- 

 latis lateralibus 1-jugim mediane 2-jugim 2-pinnatisectis, lobis oblongis 

 incisis; floribus falcatis calcare recurvo abrupte adscendenü petalis later- 

 alibus aequilongo, labiis ecristatis postico margine vix ampliato apice 

 subacuto. 



Persia : Gme/m (Hb. Willd., n«» 12916)! Flos 18 mm. longus; calcar 

 9 mm. longum; petala lateralia 9 mm. longa. 



The above diagnosis of genuine C. persica is based on a drawing of 

 the original spécimen of Hb. Willd. n° 12916, very kindly made by Pro- 

 fessor Urban, and from a single flower belonging to this spécimen, 

 very kindly lent from the Royal Herbarium, Berlin. The original descrip- 

 tion by Chamisso and Schlechtendal gives three opposite long-petioled 

 leaves; the drawing however shows ihat there is something wrong in 

 this description, for it indicates four such leaves, not three. In any case 

 thèse are only segments, with long petiolules, of two opposite leaves one 

 of which obviously has lost or has never possessed one of the lateral seg- 

 ments, the other being in turn defectivein wanting the central segment. 

 At the point where this leaf joins the stem the drawing indicates a small 

 projection which may be the rudiment or the remains of this median 

 segment but which appears more like an abortive axillary raceme. The 

 sepals are not described by Chamisso and Schlechtendal but on the 

 flower examined by me one still remained. In others respects the de- 

 scription in Linnœa^ though brief, is very good and brings ont extremely 

 well the fact that this plant is one of the group of forms for which the 

 generic name Cryptoceras was proposed by Schott and Kotschy. 



About three yeass ago, when engaged in arranging the undigested 

 material of the genus Corydalis preserved in the herbaria of Calcutta, 



