3,j4: the J0UR>'AL of BOTAXr 



usual habitat is in open woodlands upon a siliceous formation ; it is a 

 common and bj no means a thinl}^ distributed plant on the sparsely 

 wooded portions of the banks of the river Barle. It occasionally 

 occurs on roadside banks where these are damp or shaded by a wall, 

 as at Exford, or by overhanging trees as at Xettlecombe. Its occur- 

 rence in other situations is, as Mr. Tliompson remarks, almost 

 certainly due to the agency of man. During this summer I found it 

 in a recently cleared woodland, in a district where I had not previously 

 noted it. Its occurrence upon limestone soils must be looked upon 

 with suspicion ; it is not a deep-rooted plant and the soil may, as 

 Mr. Woodraffe- Peacock says (p. 225), be "acid sandy above, or the 

 upper root-soil is neutral from endless rain-wash and plant-decay." 

 Wrington Warren, where Mr. Thompson notes its occurrence, is an 

 example of a ''calcareous heath"' where many lime-hating bryophytes 

 are abundant. In July of this year I found it growing abundantly 

 in a fallow corn- held near Kayleigh's Cross on the Brendon Hills, 

 where most of the plants associated with it suggest a calcareous sub- 

 stratum, the floristic composition of the field being very similar to that 

 on White Lias pastures. Of the chief plants noted the following, 

 besides the Hij peri cum, were abundant: Geranium columhinum, Shei'- 

 ardiaarverisis, Th y mus Serpyllum, Rumex Acefosella ; Ononis repens, 

 Alchemilla arvensis, Filago germanica, Euplirasia rosthovia7ia, 

 E. ciirfa, Barfsia Odontites, Calamintlia arvensis, Plantago lanceo- 

 lafa, and Aira carifopJiyllea were occasional. The abundance of 

 H. humifasum and B. Acetosella amidst such company presents an 

 ecological problem, the solution of which may lie in the superficial 

 <listribution of humus over a calcareous substratum ; the time 

 available was iuf-ufficient for a thorough examination of the geological, 

 })hysical, and chemical data. The mosses noted were not characteristic 

 of limestone. — W. Watson. 



Argtle Records (p. 322). With the exception of Centuncnhis, 

 all the plants mentioned are already on record for v.c. 98 — Fotamo- 

 geton perfoliatus (Macvicar) in Ann. Scot, N. H. 1899, 40 and the 

 remainder by Prof. King in Ewing's Glasgow Catalogue^ 1899. — 

 €. E. SALMO^^ 



REVIEWS, 



The Enqlish Rock-Garden, By Regi:nald Farrer. 2 vols. 4to, 

 cloth, pp. Ixiv, 504, viii. 524, 102 plates. T. C. & E. C. Jack, 

 London and Edinburgh. Price £3 35.. net. 



These handsome volumes — well printed on good paper, illustrated 

 by about two hundred admirable reproductions from photographs 

 (there are two figures on nearly every plate), and suitably bound, are 

 in every way a credit to the publishei-s. The author, Mr. Reginald 

 Farrer, has long been known as an authority on Rock-Gardens, on 

 which he has already published more than one book, and which he 

 has enriched by the results of his travels. The present work, he tells 

 •us, "was written in 1913 and corrected for press in China during the 

 winter of 1914 " ; its appearance was delayed by " the exigencies of 



