37 



bergs is not very probable; all the deposits have originated 

 from the so called North bay in the floating ice, which during 

 summer forms a free passage from the open sea towards 

 the East -coast of Greenland. Generally speaking we may 

 suppose, that, in places where there is little ice during 

 the summer, less material will be deposited by the ice 

 at the bottom of the sea, than where there is closely 

 packed ice. Consequently throughout the region herein des- 

 cribed, the ice will be of no great importance as determining 

 the constitution of the samples. It might be imagined however 

 that the melting of the ice would be sufficient to explain the 

 great number of coarser ingredients found in these samples, 

 if the amount of finer material conveyed from the land is less 

 in proportion. We cannot so easily understand the reason 

 why the outermost samples are considerably coarser than the 

 innermost ones, for, as far as we can gather from all accounts, 

 there is no very great difference in the number of icebergs in 

 the outer and inner part of the tract. 



Assuming then that the coarser ingredients of the three 

 first samples originated from the rocks at the bottom of the 

 sea, here we have again two possibilities, namely that they 

 have been derived either from solid, or from loose rocks. The 

 first supposition is not very probable; in the first place we 

 cannot thereby explain the circumstance that the samples be- 

 come finer nearer land. If the bottom of the sea has ever 

 been above sea-level, this must have been the case with the 

 whole tract, and many submarine rocks must then have been 

 found also nearer land , and would hardly have been entirely 

 covered by later deposits. Moreover we always notice that, 

 where the material of the samples originates from solid rocks of 

 the sea- floor, it will always be of a rather homogeneous nature 

 in each individual sample, and will often vary very much from 

 one sample to another, but this is not the case here as will 

 be proved later. 



