GERANIUM PURPUREUM VILL. AND G. ROBERTIANUM L. : II 271 



G. robertianum var. genuinum Syme (1864), Eng. Bot., ed. 3, 2, t. 305, 203. 

 G. lohertianum var. graniticarum Martr. Don. (1864), Florule du Tarn, 126. 

 G. robertianum var. leucanthemum Dum. & var. laciniatum Beckh. ex Wilms. & 



Beckhaus (1879), 7 Jahresber. Westfal Prov.-Ver., 1878, 178. 

 G. robertianum var. typicum Fiori & Paoletti (1896-1908), Fl. anal, d'ltalia. 

 G. robertianum var. crassicaule Rouy (1897), Fl. France, 4, 95. ' 

 G. robertianum var. umbraticum Westerlund (1904), Bot. Notis., 1904, 14. 

 G. robertianum subsp. eu-robertianum Briq. ex Knuth (1912), Pflanzenr.,4, 129, 64-5. 



Gams in Hegi (1924j, Fl. Mittel-Eur., 4 (3), 1714. 

 G. robertianum var. trilobatum Takeda (1916), Rep. Bot. Soc. & E. C, 4, 191. 

 G. robertianum L. var. robertianum sec. Warburg in Clapham, Tutin & Warburg 



(1952), F/. Br. Is., 390. 



This subspecies is distributed throughout the British Isles, being commonest in 

 partial shade in areas of base-rich soils at elevations up to 2,300 ft., well above the tree- 

 line. It develops an ascending habit, at least in cultivation, where most plants behave 

 as biennials. In their first season, the rosette of leaves is rather dense and mound-forming 

 with long petioles. Stems and petioles are generally of a dull but deep red colour and 

 bear eglandular hairs especially near the base and glandular hairs in the inflorescence 

 and especially on the sepals. Leaflet- segments are rather broad. Petals are at least 

 twice as long as the sepals (with flowers usually between 1-2 and 1-7 cms. in diameter). 

 Anthers vary in colour from orange to dark purple while the fruits are more or less hairy 

 with a pattern of well-spaced ridges and reticulations on the back (Fig. 1). 



Fig. 1. Dorsal view of fruit of G. robertianum L. subsp. rohertianum from Cheddar Gorge, N. Som. (X 20). 



A great many varieties, forms, etc. have been described by those who have only 

 observed the plants in the wild. As a result of the cultivation experiments it has been 

 decided that, of those listed above in synonymy, even the forms breeding true have no 

 separate ecological significance and they are, consequently, retained within the subspecies. 

 There is an interesting contrast between Wilmott's' (1921) belief in a significant correlation 

 between glabrousness and growth on limestone in Britain and the observations of French 



