279 



than normal moss. The natural conclusion from this 

 observation is that the protoplasmic changes induced by 

 desiccation must be the same as those induced by cold. 



Irmscher, furthermore, repeated with moss the exjjeri- 

 ments of Molisch on the difference between material 

 frozen in water and that frozen in air. Moss partially 

 immersed in water was exposed for 3 days to — 5° to 

 -6\ The submerged portion, all surrounded by ice, 

 remained alive, while the tips which projected out were 

 killed. The more complete dehydration in air is as- 

 sumed to be responsible for the injurious effect. 



Cameron and Broivulee (1913), discussing experiments 

 in which frogs were killed after having been frozen by 

 an exposure of several hours at or below the freezing 

 point, attribute death to the excessive concentration of 

 the cell contents caused by dehydration. (They note that 

 such concentrated cell contents should not freeze.) 



In order to clarify the problem of the correlation 

 between ice formation and death in plants, Maximov 

 ( Wl'f) undertook to study: 1. The death temperature as 

 related to the freezing point ; 2. The infiuence on death of 

 the time the material is left at a given freezing temper- 

 ature; 3. The relation between the quantity of ice formed 

 at a given time and temperature and the occurrence of 

 death. Using potato tissue in which the vitality of the 

 cells was judged by the plasmolysis test, he found that 

 the tissues were completely killed in a short time at 

 -2.66 while almost all the cells were alive at -1.96° 

 and -2.11°. Furthermore, a piece of potato left to freeze 

 for 10 minutes between 0° and - 1.82° was alive, wdiile 

 one left for about 3 hours at the same temperatures was 

 practically all dead. 



These and other experiments confirmed his views that 

 death progresses step by step with ice formation and, 

 consequently, with water withdrawal. So this author 

 accepts, in general, the dehydration theory of ^liiller- 

 Thurgau and Molisch, but modifies it by suggesting that 



