1C II. (,. -IMMnNS. [SEC. ARCT. EXP. FHA.M 



of this di-cn--ioii. viz., marine currents am] floating in-. The cause for 

 excluding tin- a- a factor in the stocking of the islands here in question, 

 lies partly in the existence of a high ice-foot round most parts of their 

 -hore>. Castle Island was entirely encircled hy an icefoot which appeared 

 never to disappear, and, even had some part of the ice-foot along the 

 shore of Devil'- Isle hern washed away at the time of our visit, I 

 think that no single species has reached thither with the help of the 

 water. The influence of marine currents in the dispersal of plants has 

 certainly heen often much over-valued, and I can only agree with ERNST' 

 who writes: ,.lt has long heen known that only a comparatively small 

 proportion of plants are capable of extending the area of their distri- 

 hulion hy this means. A comparison of island floras has shown that it 



is exclusively strand plants which have seeds and fruits posse-- 



.-ing the necessary adaptations for this method of dispersal hy ocean- 

 currents, that is which are capahle of floating for weeks or months on 

 (-water, without losing the power of germination" (1. c., p. 5). 



But here we have not a single strand plant, and, as a rule, th> j 

 arctic lands are rather poor in halophytic plants which might stand a 

 journey in salt water. The floating ice, of course, may sometimes carr\ 

 seed- and fragments of plants I have occasionally seen blocks of ice 

 from the tidal crack, laden with masses of vegetable matter but this 

 does not prove that plants can in fact immigrate by that means of 

 conveyance; for a short drift within a fjord or over a strait, it may 

 perhaps sometimes be of use, if the ice-block takes the shore again 

 before the plant fragments are blown into the water or welted through: 

 but. as a transport over wide distances it is certainly not serviceable, as 

 the vegetable matter will be imbedded in the ice and will be unable to 

 come farther inland before being immersed in salt water. 



It may also be mentioned that both islands were formerly visited 

 bv man. I am not, indeed, inclined to attribute any influence for the 



w */ 



transporting of plants to these visits, but where people have been, one 

 ha- always the possibility of human influence to reckon with. The 

 indication of human visitors to these islands consisted especially in a 

 sort of .shelters, built for the eider-ducks to place their nests in. Now 

 such shelters are built in countries where the eider-duck is protected 

 lor the collecting of down; bill it is not known that the Kskimo have 

 done any such thing anywhere else, and this region has certainly never 

 had any other human inhabitants. Perhaps the shelters may be attri- 



: i ERNST, A.. Th.- N.-w Finn, of the Volcanic Island of Krakatau. Caml.ri.lf,'.- 11KJ8. 



