46 



depth is not especially great, viz. 226 Danish fathoms; the 

 sample No. 23, which was obtained near if, was at a depth of 

 294 Danish fathoms, and forms on the whole a transition 

 between No. 22 and 24, for which reason it has not been men- 

 tioned further. In order to bring about this remarkable con- 

 stitution all the different physical circumstances of the place 

 must have acted in concert to prevent the deposition of coarser 

 ingredients; but as most of these circumstances are not very 

 well known, it is impossible to account for the phenomenon 

 definitely. We can only say that the melting of icebergs must 

 have been almost out of the question here ; in summer they 

 must therefore always pass outside this place. Moreover, all 

 the directions of the currents must constantly be such that they 

 cannot convey hither material from the rather near-lying tracts 

 of land. Van Dyk Rock and Canning Land. If just now 

 and again a somewhat powerful current came from one of these 

 places, much of the material formed by the coast erosion may 

 have been conveyed to the deposit, and it would thereby con- 

 tain a larger amount of sand. However it is hardly possible 

 that these causes alone would be able to produce a constitu- 

 tion like that of the sample, a very great amount of clay must 

 also have been conveyed into the sea at this place. The 

 colour of the sample, which is pure grey, indicates that it 

 cannot have been deposited very slowly. It is impossible to 

 say for certain if a very great amount of clay can be conveyed 

 out of Davys Sund, as most of the natural conditions of this 

 territory are very little known. In this connection it will be 

 of great importance whether the salinity of the water in the 

 bay is perceptibly less than in the sea; this may in itself be 

 the cause of the deposition of an especially large amount of 

 clay outside a large bay like this one. 



By proceeding from No. 24 to No. 25, one of these sudden, 

 rather inexplicable changes in the constitution of the samples 

 again occurs, to which we have already drawn the attention 



