241 



So they conclude, paradoxical as it may seem, that sub- 

 cooling is dangerous while freezing is protective. A few 

 of the arguments of the main proponent of this theory, 

 ]\Iez (11)05), will suffice to illustrate it and to show how 

 it has been applied to the most varied plK'iiomeua of 

 nature. 



According to this author, the tissues of the stem of 

 Impatiois were killed sooner when freezing was preceded 

 by subcooling than when it was not. But in these experi- 

 ments the quantities of ice formed in the two cases were 

 not compared, a factor which has been shown to be of 

 fundamental importance in causing death. 



Mez then applies his theory to the supposedly observed 

 fact that trees and shrubs are more injured in "stagnant" 

 air than in air agitated by winds. Shaking would, in the 

 latter case, prevent subcooling which is assumed to be 

 the damaging agent. 



The frost resistance acquired by ''hardened" plants is 

 attributed by this author to the prevention of subcooling 

 by oils formed during hardening. 



The fact mentioned by Sachs and Molisch that a sprout 

 of Tradescautia, exposed to 5°, half in water and half 

 in air, had the latter part killed and the former un- 

 harmed, is explained by Mez on the assumption that the 

 part frozen in air was previously subcooled to a greater 

 extent. 



II. ACTION OF EXTREME COLD 



While injury and death can result from cold without 

 ice formation at near-zero temperatures, it seems that, in 

 general, extremely low temperatures are harmless if ice is 

 not formed. So, instead of discussing the mechanism of 

 injury by cold in this section, we shall consider the cause 

 of the innocuousness of xevy low temperatures. 



1. Observations. The organisms which resist the tem- 

 peratures of liquefaction of "permanent" gases can be di- 

 vided into two groups: 1. Those which support drying 



