70 THE BRITISH WILLOWS 



Borrer's plate and description in having pubescent ovaries, and 

 is rather typical S. phylicifolia. 



Three more of these named forms have the ovaries glabrous 

 below but pubescent in the upper half, a condition which indicates 

 descent from species with glabrous and species with pubescent 

 ovaries. These are S. laxiflora (G. Anders.) Borrer, E. B. S. 

 No. 2749, with obovate blades, S. propinqua Borrer, E. B. S., 

 No. 2729, with oval-oblong blades narrowed to both ends, and 

 S. -tetrapla (Walker) Sm. Engl. Fl. iv. 177, with narrow obovate- 

 lanceolate blades ; all of which are best placed among the hybrids 

 of S. Andersoniana and S. phylicifolia. 



The status of S. Dicksoniana Sm. has long been a problem. 

 B. White suggested S. arbuscula x phylicifolia as a likely solution 

 (/. c. 399, 412). Smith himself connected it with S. arbuscula by 

 its catkins ; but his S. arbuscula was a S. repens rosmarinifolia 

 form. It is possible that the sheet in Smith's herbarium may 

 solve the problem. This sheet labelled " Mr. Crowe's garden — 

 plant sent by Mr. Dickson — S. myrtilloides El. Brit." — then later 

 " S. Dicksoniana Engl. Bot. t. 1390," contains four foliage speci- 

 mens and one of April catkins. The foliage is S. phylicifolia, 

 with no suggestion of S. arbuscula in it and not much of S. repens, 

 if any. The flowering specimen is a S. repens rosmarinifolia form! 

 Is it possible that Smith, not suspecting any mistake or mixture 

 of specimens on Crowe's part, accepted the catkins and foliage, 

 said to be taken from the "plant sent by Mr. Dickson," as one 

 species, which he named S. Dicksoniana in honour of its finder ? 

 and that the S. phylicifolia plant was passed on through Borrer 

 (Leefe, Sal. exs. i. 12) as the plant of Smith? The catkins in 

 Leefe's specimens are not much different from S. phylicifolia type, 

 and bear no resemblance to a S. repens rosmarinifolia form. 



S. phylicifolia is found in similar situations to S. Andersoniana, 

 from West Yorkshire and North Lancashire northwards to the 

 Orkneys. Usually in hilly or mountainous regions, it is found at 

 quite low levels in the North of Scotland, and ascends to 2200 ft. 

 in the Highlands, but is not common on mountain rocks. In 

 Ireland it was formerly known from Mayo to Donegal, London- 

 derry, and Antrim ; seen in 1885 on Ben Bulben, near Sligo, and 

 in Westmeath (planted). Mountain regions of Central Europe, 

 Scandinavia, Russia ; in Asia, Siberia, Kamschatka, N. China. 



