1914-15.] BOTANICAL SOCIETY OF EDIXBUKGH 333 



suggests that it comes from limestone, and indeed it may 

 be regarded as on the way to the form which I have 

 described as P. parva. Balf. fil. All the features that give 

 this impression do not come out in the picture — the glaucous 

 hue of the plant for one. The short petioled leaves are 

 here more rounded and cuneate. and long hairs are found 

 on petioles, veins of under surface of leaf, scape, bracts. 

 and pedicels, and there is a beard at base of the calyx. 

 The umbel is few-flowered 1 2-5 ). The pedicels are short — 

 under a centimeter long. The calyx has short lobes, often 

 mere undulations at top of tube with some projecting 

 denticules. 



In this plant we have the form diagnosed by Franchet 

 as P. Listeri. King. var. rotu/ndifolia, Franchet. On 

 Plates L and LI are shown two sheets of Franchet's 

 variety for comparison from the Paris Herbarium. I am 

 indebted to M. Lecomte, Director of the Botanical Depart- 

 ment, Jardin des Plantes, Paris, for the loan of the sheets, 

 which have been photographed. These sheets bear 

 specimens of Delavay's collecting under Xo. 307 and 

 No. 845 — the specimen cited by Franchet as types of 

 his variety (see p. 313). 



I give these illustrations because of some confusion that 

 has been introduced into the identifications, and of this I 

 must now say something. 



Petitmengin wrote 1 about a species which he named 

 P. begoniiformis. Petitm.. and which he founded upon a 

 specimen in the Paris Herbarium having the same ticket as 

 that of the specimen named P. List* ri, King. var. (c) glab- 

 rescens, by Franchet. Of his species he says : " Near P. 

 obconica, Hance, and P. Listeri, King," and only gives the 

 following differential, not determinate diagnosis : — A plant 

 of glabrous aspect covered by very short hairs of which the 

 cell membrane is reticulate, whilst in P. obconica, Hance. 

 the plant is velvety and the long hairs have no reticulations 

 on the wall; in P. Listeri, King, the hairs are elongated as 

 in P. obconica. Hance, and have the same constructions and 

 the same wall pattern as in P. begoniiformis, Petitm. 

 It further differs from P. obconica, Hance, and P. Listeri. 

 King, in having a rhizome, fragile, covered with the im- 

 1 Bull. Soc. Sc. de Nancy, ser. 3, viii (1907), 11. 



TRANS. BOT. SOC. EDI>". VOL. XXVI. 2~i 



