APPENDIX. 565 



arrows, while the interrupted blue arrows show the course taken by the 

 Gottland boulders that went to south-east and south. It is hardly 

 possible that all these erratics could have been transported by one and the 

 same ice-flow ; they are more likely, as Mr. Helland has remarked, to 

 have travelled at different times. When the mer de glace attained its 

 maximum development we know that it reached down to Silesia and 

 Saxony, and at that time the stones carried forward with the moraine 

 profonde from the Baltic provinces of Russia would spread towards S.E., S., 

 and S.S.W. In like manner boulders from Gottland would have a tend- 

 ency to move to S.S.E., S., and S.W. But when the mer de glace had 

 become much reduced, and no longer flowed so far south, the lower strata 

 of the ice in the Baltic would be to a great extent confined to that hollow, 

 and hence the bottom-moraine would tend to be pressed and rolled for- 

 ward down the trough towards the Kattegat and the North Sea. In this 

 manner fragments detached from the rocks of Esthonia and the islands in 

 the Baltic might well come to be scattered through the drift deposits of 

 Hamburg, Groningen, and Denmark. And that this is not a mere un- 

 supported conjecture is shown by the fact that the stria? in Gottland, in 

 the southern extremity of Sweden, and in Zealand, clearly evince that the 

 ice has flowed in different directions. Thus in Gottland most of the strise 

 point to S.W., but another set of glacial groovings goes towards S. and 

 S.S.E. In the south of Sweden, again, while one series of strise indicates 

 a glaciation in an approximately southern direction, another set proves an 

 ice-flow towards S.W. At Faxo the direction varies from N.W. to W. and 

 S.W. 



But even at the period of maximum glaciation, considerable oscilla- 

 tions in the direction of the mer de glace may have been induced by varia- 

 tions in the thickness of the ice itself. If the precipitation of snow were 

 to become abnormally great in some particular region, so as to give rise 

 to a local thickening of the ice-sheet, this of itself would tend to modify 

 the direction of the ice-flow, and so bring about a corresponding modifica- 

 tion in the trend of the stones and rubbish travelling forward at the 

 bottom. And such local changes, being repeated at different times and in 

 different areas, might eventually give rise to some of the cases of abnormal 

 distribution of erratic blocks referred to above. But oscillations of this 

 kind are not required to account for the fact that rock-fragments, detached 

 from some particular district of small extent, are often distributed over a 

 very much wider area to the south. A glance at the map will show that 

 the mer de glace, as it flowed on towards its terminal line, gradually spread 

 itself over a wider and wider area. And this being so, the same must 

 have been the case with its bottom-moraine. It is not surprising, there- 

 fore, that stones derived from Gottland should have been distributed by 

 one and the same ice-sheet over a considerable area to S., S.W., and S.E. of 

 that island. 



There is yet another cause which may have played no unimportant 

 part in the distribution of erratics under the ice-sheet. I have endea- 

 voured to show (see page 239) that subglacial rivers must have existed, and 

 that these in Northern Germany would follow the general slope of the 

 land, and would thus often flow in directions opposed to the course of the 

 mer de glace. We can hardly doubt that quantities of morainic debris 



