8 Taylor, A Scientist in the Antarctic. [vd.'"^xxxvi. 



At Mount Suess, behind Granite Harbour, Debenham found 

 some small plates in the sandstone. We diligently searched 

 the locality, and found numbers of these plates — some bluish, 

 and burnished almost like beetle elytra. They have been 

 assigned to primitive armour-clad fish, and so determine the 

 Beacon sandstone hereabouts as of Devonian age. Far to the 

 south Wilson discovered well-preserved Glossopteris leaves, so 

 that there the sediments and associated coal are akin to our 

 Sydney coal-field. Here, also, Wright found a fine specimen 

 of the primitive Cambrian " coral," Archeocyathincv, in the 

 Beardmore Moraine — near where Shackleton had also found 

 relics of this fauna. In the far north Priestley also added 

 considerably to the Permian ifora of East Antarctica, so that 

 the expedition was very successful in fossil fields. 



The problem which interested me most was the evolution of 

 a glacial landscape. Research in the Alps and elsewhere has 

 shown that bygone glaciers have carved out great valleys and 

 impressed many peculiar features on alpine scenery. But 

 many problems are still unanswered. Of these the chief are 

 the origin of the cirque (or " armchair ") valleys, and the actual 

 mechanism of glacial erosion. In the great scarp bounding 

 the west of MacMurdo Sound is a series of unrivalled cirques. 

 The Walcott Cirque is twelve miles wide, with a rear wall 

 10,000 feet high, and a small glacier only a few miles wide 

 occupies this great hollow. A complete series in different 

 stages of evolution were mapped along the coast from Mount 

 Morning to Mount Marston. These examples have led to 

 what is, I believe, a somewhat novel theory of land erosion 

 known as the " Palimpsest theory." I believe that the chief 

 carving of the earth's surface in an Ice Age is done, not by 

 glaciers, but by the action of King Frost. The gradual cooling 

 and gradual warming at the onset and waning of the Ice Age 

 extends through much longer periods than the age of maximum 

 glacier development. In these lengthy periods "sapping" or 

 frost-erosion is paramount. At this time most of the cirque 

 valleys so characteristic of alpine scenery are cut out by a 

 sapping process, too complicated to describe here. As the 

 ice-fields increase and the glaciers pour down into the valleys, 

 they carry ice-erosion deeper into the crust, but they often 

 only partly obliterate the earlier erosion by sapping and frost 

 action. Thus, the earlier landscape shows dimly — much as 

 does the earlier writing in the Greek palimpsest. Most geo- 

 logists have allowed the later erosion to engage their attention 

 too often at the expense of the earlier and (in my opinion) even 

 more important erosion by sapping. 



Many of the less fundamental features of ice erosion are most 

 interesting. When a glacier reaches the sea ice it buckles the 



