VARIATION IN PEPLIS PORTULA L. 



87 



the line joining places with a hygrothermic value of about 70 and those areas in which 

 the two varieties are represented in approximately equal proportions ; in Ireland, however, 

 no such correlation is perceptible, though little can be argued from this owing to the small 

 amount of material seen from the eastern part of that country. 



It would be premature, in view of the so far scanty evidence adduced, to make too 

 emphatic an assertion about the nature of this geographical trend. It' does, however, 

 appear that a cline, albeit a rather ill-defined one, is to some extent recognisable. This 

 cline is remarkable for the fact that one extreme, var. longidentata, has a " Lusitanian " 

 type of distribution. The direction of the cline - or topocline, as it should perhaps be 

 more accurately termed - is also interesting. Most topoclines so far recognised in the 

 British Flora seem to extend from north to south. The only other recorded examples of a 

 true east- west cline are apparently those exhibited by Ulmus stricta Lindl. (Melville, 1939, 

 1944, 1948) and by Orchis fuchsii Druce subsp. hehridensis (Wilmott) Glapham (Harrison, 

 1949, 1952). 



Material from the following herbaria has been examined : 



Hb. Univ. Birmingham (B), Hb. Univ. Cambridge (C), Hb. Carlisle Museum (CI), 

 Hb. Royal Botanic Gardens, Kew (K), Hb. Manchester Museum (M), Hb. British Museum 

 (Natural History) (N), Hb. Univ. Oxford (O), Hb. N. Douglas Simpson, Bournemouth 

 (S), Hb. National Museum of Wales, Cardiff (W). 



The extra-British distribution will be considered first, 

 var. longidentata : 



FRANCE : Banyuls-de-mer. Roussillon, 1825. M. Petit (K). 

 SPAIN : Manzanares, 1841, G. F. Renter (N). 



PORTUGAL : Manteigas, 1881, /. Daveau (C) : Gollega. 1886, A. R. da Cunha (N) : Near Oporto, 1889, 



R. P. Murray (N). 

 ALGERIA : near Lake Houbera, La Calle, 1841-4, /. Gay (K). 



AZORES : Caldera, Fayal, 1842, H. C. Watson (C) and 1929, T. G. Tutin & E. F. Warburg (C). 



Rouy & Camus (1901) record this variety as occurring " here and there " in the 

 departments of France nearest the Pyrenees. In Spain it has been recorded by Knerskon 

 (1874) from Escorial and by Willkomm (1893) from the Sierra de Palma; it is probably 

 more plentiful than the sparsity of the records suggests. Wolley-Dod (1914) noted 

 that it had been recorded from Gibraltar, although his only gathering belonged to var. 

 portula. In Algeria (Battandier & Trabut, 1890, 1902) it is said to be " frequent in the 

 east," while Jahandiez & Maire (1931) report it as more frequent than the type in Morocco. 

 In the Azores Trelease (1897) records it on Flores, Fayal, Pico, Terceira and San Miguel. 

 All these authors, however, do not distinguish between the extreme type and the probably 

 equally frequent intermediates. I have seen material of the latter from : — 



SPAIN : near Madrid, 1844, M. Willkomm (K) : La Granja, Sierra de Guadarrama, 1858, E. Boissier & 



G. F. Reuter as P. longidentata (K). 

 PORTUGAL : Coimbra, 1848, F. Welwitsch (C) : near Manteigas, 1881, /. Daveau (N). 



Var. portula is the only form outside Macaronesia, North Africa and the Iberian 

 Peninsula. Material of this has been seen from most of Europe, including Italy, Sicily, 

 Sardinia, Corsica, many parts of France (including La Vendee) and a single gathering 

 from Pico in the Azores. 



British Isles 



A considerable amount of material lacked ripe capsules and could not be determined 

 satisfactorily. In the far north the species tends to become aquatic and evidently does 

 not fruit readily. 



