CHAP. Lxvi ADAPTATION OF SPECIES 253 



occur in Lloydia serotina, and Chamaeorchis alpina. On the Andine 

 fell-fields C" piinas ') and in Tibet many plants with subterranean organs 

 storing reserve-food are however stated to occur. The vast majority 

 of dicotylous plants possess a multicipital stout tap-root but produce no 

 adventitious roots ; as examples, may be cited Silene acauUs, species of 

 Arenaria, Draba, Dryas, and Saxifraga oppositifolia. 



Rarer are herbs with horizontal epigeous or hypogeous shoots that 

 strike root, and undershrubs, such as the small arctic willows, with 

 subterranean shoots. 



4. Nanism. Extremely characteristic is the prevaihng nanism, 

 which is caused by the factors mentioned in Chapter XLV as checking 

 growth. The foliage-leaves are swa//, and in many species assume rounded- 

 off, more or less entire shapes ; even among mosses the leaves are shorter 

 and broader than in individuals of the same species in other habitats. 

 On the other hand the leaves are linear in some of the plants, which 

 consequently become moss-like, as is the case with certain species of 

 Saxifraga and Sagina. The vegetative shoots are short, and have short 

 internodes, and often rosulate leaves ; whereas the flowering axes are 

 frequently more or less scape-like and bear small, bract-like leaves. 

 Alpine species therefore differ in habit often markedly, from allied or 

 parallel species in the lowlands : for instance, Artemisia nana from 

 A. campestris, Aster alpinus from A. Amellus.^ Yet the shoots are 

 sometimes elongated, though prostrate and closely apphed to the ground. 

 Woody species have bent, curved, and twisted shoots, which often trail 

 along the ground, as in Betula nana, Juniperus, Empetrum, Dryas octo- 

 petala, Loiseleuria procumbens, or are entirely subterranean so that only 

 the short tips of the branches project, as in Sahx herbacea (' turf-shrubs ').- 



5. Size of flower. Very often one can see that this nanism concerns 

 only the assimilatory shoots, and that flower and fruit are of the same size 

 on mountain and lowland.'^ When it is asserted that flowers are larger, 

 this is a subjective impression (possibly aroused by the small size of the 

 vegetative organs) and is not supported by measurements. 



6. Lie of leaves may be different from that assumed on the same 

 species in another habitat ; they become more erect, appressed, and 

 concave in Juniperus and Lycopodium,'* and are always directed upwards 

 or even vertical in some species, such as Ottoa oenanthoides and Crantzia 

 lineata which have juniper-like leaves. 



7. Cushion-like and tufted plants.^ Branching is often very close ; 

 this is due partly to shortness of the stems and internodes, and partly to 

 desiccating wmds, which kill the youngest shoot-tips and call forth an 

 irregular and vigorous development of branches.^ A part in this may 

 also be played by the fact that the flowers are terminal in many species.' 

 In this way many species acquire a very low, dense, convex, often hemi- 

 spherical, tufted or cushion-like form, which is characteristic, not only 

 of Spermophyta, but also of mosses, and is very obvious when one com- 

 pares subglacial species with allied or parallel species growing in lowlands. 

 The density of the tufts is increased by the circumstance that old dead 

 parts, including leaves, remain attached \A'ithout undergoing decay." 



' Bonnier, 18906. C. Schroter, 1904; cp. p. 26; also see Lidforss, 1908. 



^ Bonnier, 18906. ' Sec figures in Warming, 1887. ' See p. 129. 



" See p. 38. ' Reiche, 1893. ' See p. 75. 



