PRIMULA. 



often seem to like : or in among the rootage of light mountain-pines, 

 which seems to have the same effect of securing drainage, and is 

 especially affected by that especially beautiful form which haunts the 

 Cottians, and there with P. marginata produces P. x Crucis — a form 

 of much more slight and tufted habit, much smaller in general 

 development, much larger and fuller and more widely honest-faced 

 in the big round bloom. These, then, are the more marked local 

 or national forms ; to save the unwary, in case they ever appear as 

 species, uncommented, in catalogues, I quote them here in order : 



P. viscosa pyrenaica (Pax). — This is evidently the finest type : 

 larger, broader leaves than those of the other types, coarsely toothed ; 

 a multitude of flowers to tho umbel. (Eastern Pyrenees.) 



P. viscosa cynoglossi folia (Widmer), smaller leaves, oval, entire, 

 or slightly toothed ; manj^-flowered umbels. This is the prevalent 

 form of the Valais, the Graian, Cottian, and Maritime Alps ; Mont 

 Cenis, Combes de Barant, Valdieri, Boreon (with P. marginata). 

 This variety seems, in my experience, usually to insist on possessing 

 rocky places, cliffs, and crevices. It readily acquiesces and luxuriates 

 in a shady aspect, as also does P. marginata. 



P. viscosa graveolens (Pax)=P. graveolens (Hegetschw. and Heer.) ; 

 the form belonging to the Western Rhaetian Alps. Smaller leaves, 

 oblong or lanceolate, narrowing straightly to their base, and more or 

 less toothed from their middle to their point ; many-flowered umbels. 

 This form belongs essentially to the Engadine, though reported from 

 as far south as the Bergamask Alps (Widmer) and as Paznaun in West 

 Tyrol (Engadine : above Samaden ; Piz Languard). As I know it 

 this development prefers to form very wide masses in full open soil 

 and in full sun. 



P. vittata. — See the note on P. secundi flora. This species, apart 

 from other differences, has more and smaller flowers to the umbels, of 

 which there are often one above another ; and the plant has the same 

 needs and habit as the other, and with the same stripes of meal down 

 the empurpled calyx — although, unless there be a considerable tangle 

 among the species, this seems much inferior to P. secundiflora of 

 nurseries as sent out by those who say they have it. 



P. x vochinensis is the second branch of P. x Deschmanni, q.v. 



P. Walshii, Craib, is a small and lovely high-aJpine rarity from 

 Tibet, which stands quite close to P. pumilio (q.v.) for needs and 

 bOter, but can easily be distinguished by its abundance of 

 glandular hairs, whereas P. pumilio is wholly bald. 



P. W'irdii. See note on P. sibirica, and also Appendix. 



P. W'mi i. aometimes seen in catalogues and lamented in lists, is 



196 



