Stkiciit, C'ockay.xK, Laixc — Mount Arrdirxmitli Disfrirf. 345 



moutonnees, as Speiiiiit has recently sliowii (1911), the ice-scratches are still 

 to be seen on the non-disintegrated rock-surface when the boulder-clay and 

 loess are removed, such rock having been preserved from Aveathering by 

 its special hardness, the softer rock having been removed by ice-action,* 



The plant-colonization of a river-bed at the present time must be very 

 similar in character to that of the Pleistocene river-valleys, while the rock- 

 faces left by the ice, where not covered by lateral moraine or l)oulder-clay, 

 would rapidly become disintegrated and extensive shingle-slips lie formed. 

 Here loess would at first play no part in soil-forming ; this would not come 

 in till the valley glacier melted, and the earliest associations would pro- 

 bably consist of the higher alpine plants, which would later on be re- 

 placed on the drier mountains by steppe-plants. A further cause which 

 would assist the spread of these latter would be th'e gradual decrease of 

 cold water from the melting cliff glaciers, and the consequently warmer 

 soil. 



As the neve and ice of the highest summits grew less and less the supply 

 of debris, owing to disintegration due to the action o*f frost, would increase, 

 while at the same time the rivers below would be cutting their deep beds 

 and forming their terraces, plant-occupation going on simultaneously, the 

 invading plants being those from close at hand. Between the new plant 

 covering and the debris, whether of disintegrated rock or of river-shingle, 

 there would be a constant battle, the plant covering nevertheless slowly 

 increasing in area. This struggle still goes on, river-fans in particular 

 showing all stages of occupation, destruction, and rejuvenescence. Algae, 

 mosses, and lichens would be the first colonizers to settle down, followed 

 by herbaceous plants and xerophytic shrubs, which together would make 

 the first associations. Trees would come later as. soil became more plentiful, 

 but they would occupy only gullies, sheltered terrace-slopes, or ground 

 where shrubs had already settled and some shelter w^as provided. 



(B.) Climate. 



Besides the physiographic factor with which is bound up the various 

 edaphic distinctions, climate has played an important part not only in the 

 distribution of species, but in the selection and evolution of growth-forms. 



As in other Canterbury river- valleys, there are here two distinct climates, 

 a wet and a dry ; or, to give them ecological designations, o jorcst and a 

 steppe climate. 



The forest climate occurs only near the source of the Rakaia, extend- 

 ing eastwards for perhaps 5 km. from the dividing range, and marking 

 by an irregular line the average distance reached by the north-west rainfall 

 (see also Cockayne, 1900, pp. 117, 118, and 130-33), the rain especially follow- 

 ing a main river-valley. When the wind is blowing a gale from the latter 

 quarter there is frequently not a drop of rain in the vicinity of Lake Heron, f 

 and transpiration conditions are at a maximvim, but from the neighbour- 

 ing heights a heavy rain-storm can be seen in progress in the Rakaia 

 Valley. 



To be sure, the steppe climate is far from being really dry, but clear 

 skies with strong insolation are frequent, and the ever-present wind would 

 demand a much higher rainjall before forest could establish itself naturally. 



* This is just the same as occurs iu Europe, &c. 



t The great mass of ^lonnt Arrowsmith also keeps back no small amount of raiu. 



