152 Transactions. 



by subsequent stream erosion, why the headwaters of the streams in the 

 valleys running south from the Cass were rapidly diverted to the main 

 river, a change which would have been accelerated by the blockage of the 

 lower parts of these valleys by morainic bars and by shingle fans, as will 

 be indicated presently. Thus it is that the Cass has captured the stream 

 issuing from the Craigieburn Mountains near the Grasmere Station, and 

 also the waters of Lake Grasmere and Lake Sarah, and numerous other 

 small streams as well. The resemblance to the conditions of the Lake 

 Stream in the Rakaia VaUey is very close indeed in this respect. The 

 presence of the morainic dam south of Lake Heron, and the overdeepening 

 of the main Rakaia to the north, has residted not only in the capture of 

 small streams in the vicinity of the outlet of the Lake Stream, but also 

 the waters of Lake Heron itself and its feeders, and, most important of 

 all, the Cameron River, all of which belong to the Ashburton basin rather 

 than to that of the Rakaia.* The parallelism between the features of both 

 localities is indeed most remarkable. 



The district furnishes excellent examples of shingle fans in all stages 

 of development ; in fact, I do not know of any locality in New Zealand 

 where a more representative series can be obtained, the country in the 

 immediate vicinity of Lake Pearson being especially prolific in the varying 

 forms of this landscape feature. A common mode of origin, excellently 

 illustrated by the district, is from the gutters or runnels which sometimes 

 seam the mountain-side for many hundred feet. These are no doubt 

 formed at times by a single heavy shower of rain, and their channels are 

 marked by raised banks and levees, and by a cone of detritus at their lower 

 end on more or less level ground — the deposition of the load of debris being 

 due in both cases to the check in the velocity of the stream, in the first 

 instance by friction with the side and the adjoining vegetation, and in 

 the latter by the lowering of the grade. The deposit at the lower end of 

 the gutter grows as material is supplied, and the ruimel enlarges its basin 

 by continued erosion of the sides. As the stream increases in volume the 

 angle of elevation of the fan diminishes and its fringe broadens, and, if 

 space is allowed, its idtimate size will depend on the volume of w^ater and 

 the supply of waste. Fans whose diameter exceeds a mile are by no means 

 uncommon in the Cass region, but they attain greater development in such 

 unconfined areas as the Canterbury Plains. The stream usually occupies 

 the highest radius of the fan, contained by levees on either side, but sooner 

 or later it breaks from this unstable position and commences building on 

 another radius. Construction along successive radii goes on steadily, and 

 the deposit of one overlaps that of another, so that the surface of the fan 

 is raised progressively. The establishment and maintenance of the plant 

 covering has a material influence on its growth, and there is evidently a 

 close connection between the condition of this covering and the stage of 

 evolution of the fan, the geological and the ecological factors acting and 

 reacting on each other. As a rule, when a fan has reached a moribund 

 condition the detrital matter is completely covered with a closed plant 

 formation. The biological station is placed on a fan which is practically 

 dead, and there are good examples of fans in a similar condition en the 

 north side of the Waimakariri, where they are covered partly with forest 



* "Th? Mount Arrowsmith District : a Study in Physiography and Plant Ecology." 

 Trans. N.Z. Inst., vol. 43, 1911, p. 341. 



