SHOET NOTES. 153 



Under the species recently published in these pages. It had to be 

 gathered rather too early, the fruit not being ripe ; some of the 

 pods show slight traces of reticulate veining, but this may be due 

 to shrinkage, and therefore merely apparent. Mr. Hanbury has 

 lately sent me specimens of the plant which I believe to be true 

 C. yroenlandica L., from The Heogs, Unst. — Edward S. Marshall. 



Pyrus latifolia Syme in E. Eoss. — In Journ. Bot. 1893, p. 231, 

 I reported P. toruuiialis from the Conan River. At that time I 

 possessed no specimens of i'. latifolia, and did not know the plant. 

 Recently, when laying in a sheet of the latter, I at once saw that 

 the Ross-shire plant was identical with it. Only one tree was met 

 with (a remarkably fine specimen in good fruit), so that its nativity 

 must be considered very doubtful, although it was not an obvious 

 introduction. — Edward S. Marshall. 



Aeabis PETRyEA Lamarck. — Lamarck describes his Arabis petrcca 

 in the Jinci/clupedie Metkodiqiie, i. 221. He says the leaves are 

 " sinuees, presqu'en lyre, a dents ou decompures obtuses, vertes, 

 glabres, et quelquefois legerement ciliees a leur base. . . . J'ai trouve 

 cette plante en Auvergne, sur les pentes seches des montagnes." 

 He gives as synonyms Canldiiiine petrtBa ; cmnhrlca, nasturtii facie, 

 Dill. Elth. 70, tab. Gl, f. 71, and Cardainine pttrtBa L. There 

 appears to be considerable doubt whether the plant from the 

 Auvergne is identical with the plant of the Species Vlantarum, 

 which is, I believe, represented in the Linnean herbarium by 

 A. petreca yav. fceroensis of DC. Syst. ii. 230, and is figured under 

 the name Cardamine fceroensis in FL Danica, t. 1392 (not 1382, as 

 given in the Systema) ; or with the plant discovered on Snovvdon by 

 Thomas Johnson, which is A.petroia var. hispida of DC. Syst. This 

 hairy form was mistaken for Sisymbrium, arenosum by Linnaeus 

 [Flur. Suec. ed. 2, 233). It appears highly probable that the 

 plant from Auvergne described by Lamarck is a form of Sisy)iib)inm 

 arenusum, since Arabis petreca is not recorded from France in 

 Nyman's Conspectus, or in any Flora of France that I have con- 

 sulted. I thought an examination of the herbarium of Lamarck 

 might throw light on the subject. For many years his herbarium 

 was at Rostock — I believe, in the possession of Prof. Roeper — but it 

 has now been added to the herbarium in the Jardin des Plantes, 

 where I had recently an opportunity of examining it. It is kept 

 separate from the general collection there (as is the case with the 

 herbaria of Tournefort, Michaux, &c.). 1 looked through the 

 CruciJ'era, but was unable to find a specimen of Arabis petra:a. 

 I may say that the specimens of Lamarck are not mounted, but 

 are kept loose in the old covers. Very few localities are given. 

 Arabis petraa is a polymorphic species, but I should be inclined to 

 keep separate as a subspecies the A. Crantziana of Ehrhart, which 

 is figured in Reichenbach's Icoiies FL Germ, et Helv. ii. fig. 4323. 

 I have seen an authentic specimen, which has the long, narrow 

 pod, as figured there. It is more correctly called A. hispida 

 Mygina. Crantz mistook it for A. Thaliunu. This species has 

 not, so far as I am aware, been found in Britain. — G. C. Druce. 



