142 THE SCOTTISH BOTANICAL REVIEW 



insolation. The lower layers of shingle are usually moist, 

 like the crevices of rock, owing perhaps to protection from 

 evaporation. 1 Shingle banks in rivers often support a very 

 varied open association of mixed annuals and perennials, 

 vagrants and colonists from neighbouring associations. Such 

 open associations may show a mixture of plants with very 

 different requirements for water, which is often abundant at 

 a short distance beneath the surface. The older river shingles 

 in moist, exposed moorland districts accumulate acid humus 

 owing to the growth of certain species of mosses in the 

 crevices exposed to light, but in the more sheltered and 

 warmer positions they frequently become occupied by thorn 

 scrub, including species of rose, sloe, whins, and brambles, 

 with often honeysuckle, elder, broom, and others. 



In the upper courses of many of our moorland streams the 

 alluvia often consist of great stretches of shingle due to the 

 erosion and sorting by the stream of drift full of boulders and 

 stones of various dimensions. These shingles remain barren 

 if the materials are well assorted, and consequently mobile 

 during conditions of flood, but where many boulders occur 

 the intervening hollows are often filled with fine gravel and 

 sand, and the part of the bed which is deserted when the 

 stream is low forms a scattered open formation of lithophytes 

 on the boulders and sand-dwellers in the hollows. These are 

 chiefly plants that can withstand flooding and some amount of 

 scour, such as RJiacomitrium aciculare on the boulders, and 

 species of Sagina, Bryu7n, Polftrzchum,e.\.c., in the hollows. It 

 is on these shingle stretches, moreover, that certain alpines, 

 especially those which frequent springs or flushed ground, such 

 as Epilobiuni alpinum, Saxifraga stellaris, and Alcheniilla 

 alpina, may often be found several miles from the sources of 

 the streams where their centres of distribution occur. 



Shingle stretches of this nature naturally never become 

 grassland by any kind of plant succession, except at the tails 

 of the banks, which often consist largely of sand, or in old 

 deserted river loops which have become choked with fine 



^ See, however, "The Shingle Beach as a Plant Habitat," by F. W. Oliver, 

 "The New Phytologist," vol. xi. No. 3, March 1912, p. 98, where the water 

 problem of shingle beaches is discussed, and where the suggestion is made that it 

 may depend upon the formation of internal dew. 



