THE MOLECULAR FORCES IN PLANTS. 127 



air in the neighbourhood of the tuber. The whole apparatus is 

 placed in a large dish, and the investigation is carried out in a 

 cold room. We surround the bell-glass in the dish with a freezing 

 mixture (snow and salt), and then read off every five minutes the 

 position of the mercury in the thermometers. The temperature of 

 the potato gradually sinks to 3 or 4 C. Suddenly however it 

 rises again to 1 C., remains for some time fairly constant at this 

 point, and then again sinks till the temperature of the air sur- 

 rounding it, e.g. 8 C., is reached. When potatoes are exposed to 

 a temperature below C., they first of all become supercooled, 

 without any formation of ice taking place in the tissue. When 

 the maximum of supercooling has been reached, formation of ice 

 suddenly sets in, and, through the consequent liberation of heat, 

 the temperature of the potato rises to its freezing point, which is 

 situated at about 1 C. The temperature of the tubers then 

 gradually sinks to that of the surrounding medium. Other plant 

 structures behave in a similar manner. Thus, e.g., I wrapped 

 round the bulb of a mercury thermometer a strip of a Begonia 

 leaf (B. manicata), fastened it with thread, and then cooled it. 

 The maximum of supercooling lay at 4'8 C., the freezing point 

 at 0.8 C. Not until this temperature was reached did the strip 

 of leaf become discoloured (see 47). 



For exhaustive researches on the changes of temperature in 

 freezing plant structures we may employ a freezing chamber of 

 the following construction (Miiller-Thurgau). It consists of a 

 cubical wooden box, with sides 1 m. long, and with double bottom 

 and sides packed with dry sawdust. Inside this box fits exactly 

 a double- walled sheet zinc case, also open above, the inner wall of 

 which can be lifted out. To use the apparatus the outer zinc case 

 is first filled with pieces of ice to the level of the pillars supporting 

 the inner case. The inner case is then put in place, and the space 

 between the sides of the two cases is also filled with ice. If it is 

 desired to obtain in the case a temperature below C., e.g. 8 C., 

 we place on the top of the ice between the walls of the zinc case 

 a certain quantity of salt, which causes a portion of the ice to 

 melt, and so lowers the temperature. For cover we use two quad- 

 rangular trays of sheet metal, which when laid side by side cover 

 the zinc case, but do not quite reach to the upper edge of the 

 wooden box. Between the trays, which are also to be filled with 

 ice to which salt is added, is left a slit, about 2 cm. wide but 

 enlarging above, which extends across the whole case, and through 



