murphy: pexguins of south Georgia. ii,S 
colonv usuall\- selects the suininits of the windy, shelterless ridges for its 
home. \\'h\- should marine birds which lack altogether the power of 
flight, and which are at best indifferent walkers, prefer to make the 
period of propagation difRcult for themselves b}- retreating as far as 
possible from their only source of food ? The question has already been 
raised by Levick ( 1914) with reference to the great colon}- of Pygosaiis 
adclia- at Cape Adare, where many of the breeders climb to the bare crest 
of the cape and make their nests three hundred meters above sea level.* 
In the case of neither P. papiia at South Georgia nor P. adelicr at Cape 
Adare can the factor of .self- protection offer a satisfying explanation, for 
both species have only a single terrestrial enemy, the skua, to the ravages 
of which they are as su.sceptible on the highlands as on the shore. Dr. 
Levick does not explain the fact, merely .saying that it is "the result of 
their love of climbing." A consideration of the history of South 
Georgia, however, may helj) in an interpretation of the strange instinct 
which drives penguins of the genera Pygoscclis and Eudyptcs to nest 
among the hills. 
Although South Georgia is little larger than Long Lsland, X. Y., its 
glaciers are as mighty as tho.se of Spitzbergen, and there is ample 
evidence that the island was formerly completely buried by an ice-caji. 
The interior, which rises to an altitude of more than two thousand 
meters, is no longer ice-clad, excepting on the peaks, but is covered with 
an everlasting neve of the Alpine type. This consolidates at the sources 
of all the valleys to form tongues of ice, most of which extend clear into 
the sea, ending in abrujit walls. Since mo.st of the fiords have been 
carved out by former extensions of the valle\- glaciers, the coast is almost 
beachless, the few areas of low, flat land being terminal moraines or beds 
of moribund or extinct glaciers. E\"en now, with the fluctuating seasons, 
tlie glaciers .sometimes advance their fronts and flanks over con.siderable 
ground once abandoned. f It is jjrobable, however, that glaciation is on 
the wane and that an appreciable decline has taken place even since the 
\-i.sit of James Cook in 1775. Certainly Ca])tain Cook's detailed account 
of unmelting ice and the ab.sence of fresh-water streams in midsummer 
does not accord with present conditions. 
From the foregoing data it nia\- be a.ssumed that during a long ])eriod 
* The same trait is equally well illustrated at llacquarie Island hy one or more species of 
crested penguins {Eudypfes). a preliminary account of which is given in Mawson's T/ir Hnmt' of Uw 
Blizzard, London. 1915. 
