320 Transactions. — Geology. 



This \vas not the tendency with the agent which built up the 

 Timaru loess ; on the contrary, the effect was to increase the 

 unevenness. The bands of stratification curve with definite 

 relation to the drainage-lines. It is easy to understand how 

 such curves were produced, on the dust-heap theory. They 

 are the resultants of the conflicting forces of equable deposi- 

 tion, and inequable denudation by contemporary rains, the 

 latter having greater power on the slopes and surfaces near 

 the drainage-lines. It appears to me that this curved strati- 

 fication, the formation being superficial, undisturbed, and 

 resting upon a level base, is a crucial test, and settles the 

 question. 



The loess being of iEolian origin, it necessarily follows that 

 it belongs to a glacial age, for no other agent than ice could, in 

 this latitude, produce material of this kind, and uuder such 

 related conditions that the material could be spread out for 

 winds to lift and bear it away to new fields of deposit. As we 

 saw from an examination of the products of the first cold age, 

 the Canterbury Plains were principally built up long before the 

 loess period, but the seismic disturbances dming the interval 

 produced alterations in levels which we have few, if any, 

 means of fairly estimating ; and where those dust-fields were 

 cliiefly situated is a question it would be difficult to answer. 

 The building-up of shingle-fans implies the overflowing of their 

 banks by the fan-building rivers, and perhaps we should look 

 to such action for the spreading-out of the dust which was 

 swept over the higher lands by the breezes. In that case we 

 should have the whole of the Canterbury Plains, and any con- 

 temporary extension of them eastward, by changing positions, 

 as the immediate source of the dust. Against this idea, how- 

 ever, is the fact that the shingle of the fans appears to be free 

 from glacier silt, and also the fact that the fans are traceable 

 to, and appear on the whole to be coiitemporary with, the 

 moraine-dams of the mountain lakes, while the loess must 

 be older. It appears to belong to the whole of the second 

 glacial period, and principally to the earlier stages and the 

 culmination of its severity, rather than to the latest, the 

 inoraine and fan-building stage. 



As described in my paper of last year, the loess contains 

 marks of several pauses in its deposition, in bands containing 

 (a) drought veins,* the product of a dry climate ; (/?) rust- 

 granules, theprodiict of a wet climate ; (c) multitudes of birds' 

 cz'op-stones, which I shall presently suggest have an interesting 

 significance as an index of climate ; and {d) at one level certain 

 alterations of texture produced by extreme severity of climate. 

 Deposited upon areas elevated above the reach of rivers, this 



* Previouslj' described under the name " evaporation veins." 



