A PLANT THAT MELTS ICE 



35 



Still, so slight is the total quantity of heat the poor 

 little plant can evolve with all its efforts, that by 

 the time the stem is an inch or two long, the lower 

 part of the tunnel has curiously frozen over again, 

 by the process which Tyndall called " regelation," 

 and whose impor- 

 tance in glacier 

 action he so fully 

 demonstrated. 

 Inthisstage,then, f 

 the melted space fr ' 

 is no longer a 

 dome ; it assumes |E^E;E::E:;;:: :;; 

 the form of a 

 little balloon or 

 round bubble of 

 air, surrounding 

 the flower-bud. 

 At the same time, 

 the ice beneath, 

 having frozen 

 again, almost 

 touches the stem, 

 so that the bud 

 seems to occupy 

 a small, clear 



area of its own in the midst of the sheet, with 

 ice above, below, and all around it (No. 3). 

 You would say that growth under such circum- 

 stances, in almost icy-cold air, was impossible but 

 if you examine the ice-sheet at the edge of the 

 neve, you will find it studded by hundreds of such 



NO. 3. BUD, SOMEWHAT LATER, EN- 

 CLOSED IN A GLOBE OF AIR WITHIN 

 THE ICE-SHEET. 



