166 



H. G. BAKER 



(3) (4) (5) 



Figs. 3-5. Dorsal views of three fruits from the population of subsp. forsteri at Clymping, W. Sussex ( X 20), 

 showing a transition from the typical purpureum pattern (left) to a rohertianum pattern (right). 



rather more difficult. There has been some doubt as to the correct name for this sub- 

 species. Rouy (1897) describes his maritime G. purpureum Vill. var. littorale as having 

 the petals " une fois plus longs que le calice." These petals (the measurement refers to 

 the limb) are certainly large for G. purpureum and for this reason Wilmott (1921) dismisses 

 the possibility that this represents the same taxon as his var. forsteri. This could not be 

 regarded as the last word on the subject for whereas Wilmott concluded, quite unjustifiably, 

 that the "rarement presque orangees" added to the "antheres jaunes" in the description 

 of purpureum was "presumably added for the benefit of this G. Rohertianum form" 

 he admitted that he had seen no material from Rouy's type-locality (maritime shingle 

 between Cayeux and le Hourdel, dept. Somme). Actually Rouy points out that it is 

 in his var. semiglahrum that the anthers are rather orange. 



However, there is material of Rouy's var. littorale at Kew from the classic locality 

 and an examination of this sheet together with several other collections from the same 

 population (in the Paris herbarium) shows unquestionably that it consists of the nearly 

 glabrous prostrate maritime form of G. rohertianum which was first described by Babington 

 (1851) as var. maritimum. Consequently Wilmott's name var. forsteri is genuinely the 

 first to have been given to prostrate G. purpureum. 



The following stations for subsp. forsteri are represented in British herbaria : — 

 CHANNEL ISLES: 



GUERNSEY : Bordeaux Harbour, 1892, Marquand & Groves (B.M.), 1902, Cooper (B.M.), 

 1912, 1914, Barton (B.M., Camb., Ox.); St. Sampson's Bay, 1894, Gray (Camb.); Fort Cum- 

 berland, 1876, Wright (B.M.); Fort Doyle, 1892, Miller (Camb.). 

 V.c. 11. S. HANTS. : Stoke 's Bay, 1829, Borrer (K.Borr., B.M.— Syntypes), 1871, Warner (B.M.Bosw.), 

 1878, 1883, Groves (B.M.); Southsea, 1836, E. P. (B.M.); Portsmouth, n.d., Macreight (K); 

 Calshot, 1871, Streatfield (B.M., B.M.Bosw.); Hurst Castle, 1924, Turrill (K), 1928, 1929, 

 Marsden-Jones (K), 1934, Ballard & Hubbard (K). 

 13. W. SUSSEX : A pair of compact populations sampled by many collectors and variously referred 

 to as at Clymping, Middleton, Atherington or simply " between Littlehampton and Bognor," 

 type and syntypes, 1919, A. J. Wilmott (B.M.). Also an isolated plant "close up to the sand- 

 hills at Littlehampton," 1919, Wilmott (B.M.). 



