Being Comments upon the " Challenger " Reports, 4i7 



its head^ there is a depth of from 70 to 100 fathoms. Its 

 numerous islets are all flat-topped, having been planed down 

 to a common shape and level ; and these truncated islets 

 have their surfaces covered with erratics. * 



Thus we see that the island presents every indication of 

 its having been more extensively ice covered not long since, 

 and the great changes in the climate of Kerguelen's Land 

 indicated by the existence of forests of fine timber at one 

 time, and by the occurrence of an ice-sheet enveloping the 

 entire island at another, help to confirm the conclusion 

 which has been drawn from the facts of Australian geology 

 that the southern hemisphere has had its secular climatic 

 mutations. Thus a large question at issue is being gradually 

 reduced to the lesser one, of whether the temperature 

 oscillations were local, and due to local causes, or universal 

 in obedience to some less obvious and more cosmic 

 influences. 



There are some other interesting features connected with 

 this island. 



The flora of the island is poor. The ranker vegetation 

 dies out at an altitude of 300 feet, but flowering plants 

 cover the hill sides up to 1000 feet, and they exist in 

 patches where there is shelter up to 1300 feet. 



Under the protection of the herbage a few species of 

 inactive, wingless flies, gnats, and moths crawl, their habits 

 and structural modification exhibiting an interesting rela- 

 tion to their small, isolated, and boisterous habitat. The 

 higher fauna is restricted to birds and seals. The climate is 

 typically insular^ with the small annual range of 9 degs., 

 the mean summer temperature being about 45° F., and the 

 winter 36° F. Even in the depth of winter the ther- 

 mometer at the sea-level rarely falls below the freezing- 

 point, and the snow never lies on the low ground for longer 

 than two or three days.-)- 



When we remember that the weather range is perpetually 

 snow-covered at a small altitude, and that Heard Island, but 

 three hundred miles to the southwards, is glacier-covered 

 to the water's edge, we must be struck with the circum- 

 stance that the Kerguelen vegetation is perennial, and not 

 annual, and that it is not only evergreen, but that some of its 

 plants were observed by Ross to be in flower at midwinter. J 



* Challenger, 332 ; Ross, Vol. I., p. 69. + Challenger Bejports, 354. 



I Boss, Yol. 1., p. 86, and Challenger Reports, p. 355. 



