Cockayne. — Ecological Studies in Evolution. 21 



straggling mats in G. discolor Hook. f. and G. Walkeri T. Kirk, loose circular 

 cushions in G. viscosa Hook, f., and true dense cushions in C. sessiliflora 

 Hook. f. and C. argentea T. Kirk. 



Frequently the epharmony of such cushions can be seen clearly in one 

 and the same species, as in the tiny taxad Dacrydium laxifolium Hook, f., 

 which forms cushions on dry pumice at 1,200 m. altitude near Mount Rua- 

 pehu, but which growing amongst other shrubs under more mesophytic 

 conditions is frequently a straggling shrub, or when in colonies on sour 

 peaty ground merely a close turf. 



The cushion form culminates in the great amorphous masses of certain 

 species of Psychrophyton and Hoastia, which grow on alpine rocks* exposed 

 to sun, frost, and wind, or at times, in the case of R. Goyeni T. Kirk, of 

 Stewart Island, on wet peat. 



Excepting with regard to the physiologically different bryophyte cushions 

 of moors or wet forests, the cushion form is governed by strong xero- 

 phytic conditions, and the same species may thrive either in physically or 

 physiologically dry stations — e.g., Phyllachne Colensoi Berggren (Stylid.), 

 Psychrophyton Goyeni Beauverd (Compos.). 



The form under consideration occurs in the following families : Taxaceae, 

 Gramineae, Gyperaceae, Centrokpidaceae, Xuncaeeae, Portulacaceae, Caryo- 

 phyllaceae, Leguminosae, Violaceae, Thymelaeaceae, Umbelli ferae, Bora- 

 ginaceae, Scrophularinaceae, Plantaginaceae, Stylidiaceae, and Gompositae. 



Epharmonically similar cushions occur amongst different genera and 

 families in high mountains everywhere. Certain erect shrubs when wind- 

 swept become virtually cushions. 



(c.) Lianes. 



Climbing-plants have most certainly descended from non -climbing species 

 which through shade and moisture have grown upwards out of the lower 

 tiers of vegetation in a stratified association. Many transitions between 

 climbing and non-climbing plants can be observed, and these, considered 

 along with the heredity of the climbing habit and its strong differentiation, 

 afford weighty support to a belief in the heredity of epharmonic characters. 



The fern Hypolepis distans Hook. , which generally gives no hint of a pro- 

 pensity to climb, when growing alongside a support may lengthen its fronds 

 for considerably more than 1 m., though at this length they would fall but 

 for the support. On the rhachis are minute excrescences, which, though 

 certainly not adaptations for the purpose, f assist the frond to maintain 

 its position. The climbing form of Pteridium esculentum, already noted, 

 is specially interesting because of its hint at winding. So, too, with the 

 scrambling liane Lycopodium volubile Forst. f., which, gaining a thin support, 

 winds freely, the winding being in this case an hereditary characteristic. 



The case of Fuchsia Colensoi Hook, f., already mentioned, is of especial 

 moment. This is a shrub in the open, and at times a scrambling liane in 

 the forest. There can be little doubt that this latter habit is hereditary 

 to some extent, and it is possible that there may be climbing and non- 

 climbing races. This is the more likely as the " species " is considered 

 variable, and large forms are said to " almost pass into F. excorticata " 

 (Cheeseman, 1906, p. 187), which is a small tree or shrub, but never a liane. 



* Maastia pidvinaris appears to grow on shingle-slip, and not on rock, so far as I 

 have observed ; but I am also advised that at times it grows on rock. 



f Strictly speaking, there is no " purpose : ' in any adaptations, but it is often con- 

 venient to speak teleologically. 



