24 THE JOURNAL OF BOTAISTT 



from the fruiting pieces alone; but I must add that mv diagnoses 

 were greatly helped by the generous supply of material Mr. Stelfox 

 sent me, and I have greater confidence in the majority of the names 

 I have given him than in those for any collection I have yet seen. 

 The genus Rosa as represented in Britain unquestionably consti- 

 tutes our most dilticult one. On this subject I will not now spread 

 myself, but much of the doubt I often have to express on specimens 

 is due to collectors sending me the ends of flowering shoots, which 

 are almost useless. The colour of the petals is of small value, and 

 both ])riclvles and leaflets are often abnormal in shape on the ends 

 of the flowering shoots ; while the direction and persistence of the 

 sepals on the ripe fi-uit, as well as the shape of the latter, is quite 

 lost in flowering specimens. Pieces of barren shoots also are often 

 abnormal, the best specimens being those cut from the wood of tlie 

 previous year with two or three flowering branches attached. More 

 than one specimen from each bush also facilitates diagnosis. — 



A. H. WOLLEY-DOD. 



^Salsola. caefka Sparrman. This name, which is published in 

 Sparrman's Voyage to the Cape of Good Hope (17S5) does not 

 seem to have been taken up by subsequent authors. It is evidently 

 svnonymous with S. apliylla L. f. (Suppl. 173: 1781), but it may 

 be worth while to extract the passage relating to it : " There is 

 another shrub frequently found in the Carrow, which grows here 

 [Lange-dal] likewise, and is called Canna-hoscli ; whence the whole 

 tract of country hereabouts bears the name of Cannas, and not 

 Canaan'' s Land, as Mr. Mason [Masson] has called it in the Philo- 

 sophical Transactions [Ixvi. 287] ... A road between Artaquas and 

 Lange-kloof ... is called Canna's hoof/teov Canna's heights. Having 

 examined this same Canna-shrub, I found that, in strict propriety, it 

 formed a new species of Salsola ; for which reason, in my manuscript 

 descriptions of plants, I have called it Salsola caffra^foliis minutis 

 siihrotundis, carnosis, co?icavis, inihricafis. The leaves have a 

 bitter salt taste, and burned together v/ith the whole shrub produce 

 very strong ashes, excellently well ada])ted for the purpose of making- 

 soap ; for whicli reason, particular attention is paid by the CaiTow 

 farmers to the culture of this plant. In the parts of the flower, the 

 Canna-shYnh so far differs from the generical character of the Salsoltf 

 in the sixth edition of the Getiera Plantarum, inasmuch as this 

 species has a little obsolete style, with two or three brown stigmas. 

 The remaining parts of its character were, Stam. jil. hreviss. 

 Antherce cordatce, Calyx perianth, ptersistens, Gapsitla 5 valvis, 

 1 locularis, and semen 1 cochleatum, as in the Salsola, or rather 

 resembling a watch-spring coiled up " (i. 297 : I quote from the 

 second edition (1786), but believe this does not differ from the first). 

 In Flora Gapensis, v. 1. 453, the vernacular name is given as 

 " Brak Ganna." — James Bkitten. 



Cystopteeis MONTANA IN Banffshibe. In the Life of a Scotch 

 Naturalist, Thomas Edward, by Samuel Smiles, ed. 3 (1877), it is 

 recorded that from time to time Edward sent occasional natural 

 History notes to the local Banfshire Journal : in one of them 

 he reported that this beautiful Fern had been "found on Benrinnes 



