SALVIA MARQUANDII. 227 



those drawn from the shape and colour of the flower ; at any 

 rate, the latter characters should not be ignored. 



Modifications in the description of *S'. clandestina are 

 made in the later editions of Babnigton's Manual^ but they do 

 not fit the restricted plant, and it is difficult to believe he had 

 the true species before him ; nor do they agree with the 

 Guernsey plant. 



In the third edition of English Botany^ Syme, with 

 a query, identifies a plant (which is preserved at Kew) which 

 he has seen in the Borrer Herbarium, gathered in Guernsey, 

 as IS. clundestina, and this is, I think, identical with the plant 

 which I am about to describe. The figure, t. 1057, is rather 

 poor, and the colouring bad, as our plant has clear blue (beau 

 hJeii)^ not purplish flowers. Syme gives several synonyms, 

 some of which belong to the true clandestina, but none, I 

 think, to my plant. 



It may be well to state Avhat I consider to be the true 

 S. clandestina L. Fortunately there is not great difficulty in 

 this case in arriving at a conclusion. Linnaeus diagnosed it 

 {Sp. PL ed. 2, p. 36) as " 8. foliis serratis pinnatifidis rugosis- 

 simis, spica obtusa, corollis calyce angustioribus " ; he cites as 

 a synonym " Horminura sylvestre, inciso folio, cassio flore, 

 italicum. Barr. rar. 24, t. 220," and gives a detailed descrip- 

 tion. A reference to Barrelier's work (PI. per Galliani. 

 Hispaniam et Italiam ohservatce, 1714) shows that the plant 

 there described and figured differs essentially from the 

 Guernsey plant ; it is *S'. clandestina L., common in many 

 parts of Si)ain, France and Italy, which Barrelier saw on the 

 Roman Campagna. 



The identity of the true clandestina. was somewhat 

 obscured by its being represented in the Linnean Herbarium 

 by the eastern S. controversa ; hence Smith, who then had 

 the Linnean Herbarium in his possession, when he prepared 

 Sibthorp's Flora Grceca and wiote the Prodromus, described 

 and figured S. cnntrorersa as S. clandestina. L. 



It would seem probable that the excellent plates in Jordan 

 and Fourreau s Icones ad Floram Kuropoi (where several 

 Salvias, under the generic name of GaUitricliuni, are figured) 

 might have included one representing our plant, but I am 

 unable to match it : those having a concolorous corolla have a 

 very differently-shaped flower, so that only in an extremely 

 aggregate sense could they be considered to belong to the 

 same species. 



Under S. pratensis., in Corbiere's Noarelle Flore de 

 jSorniandie^ p. 453, there is described var. parviflora Lee. & 



