126 



PHYSIOLOGY OF NUTRITION. 



that not cell-sap but almost pure water has frozen out of the cells. 

 Under certain conditions, it is true, ice can form within the 

 cells of freezing plant structures, but usually the water passes 

 over into the intercellular spaces, or other cavities in the tissue, 

 and freezes there. We cut off the upper part of a large beetroot, 

 and fix it in place again with thread, after scooping out not too 

 large a hole in the lower piece. If we now for some time expose 

 the root to a temperature of say 8 C., we shall find that consider- 

 able quantities of ice accumulate in the cavity. 1 



To prove that in freezing plants the first formation of ice takes 

 place in the intercellular spaces, sections of frozen potatoes or 

 carrots, cut with a very cold knife, are laid on a well-cooled object 

 glass, and observed under the microscope while slowly thawing. 

 a We see that the ice crystals have 



formed not in, but between, the 

 cells, and that the cell rows which 

 were forced asunder by the ice, 

 naturally no longer come into com- 

 plete contact when the ice melts. 

 Therefore when plant structures 

 freeze, the cells give off water. 

 This next appears in the intercel- 

 lular spaces as ice, and as the 

 masses of ice grow, the intercellular 

 spaces also increase in size. 



It is very instructive to investi- 

 gate the changes of temperature 

 taking place within plant structures 

 during the process of freezing, as 

 was first done by Miiller-Thurgau. 

 f . I have employed for this purpose 



FIG. 43, Apparatus for investtgat- . 



ing the changes of temperature in the arrangement represented in 

 freezing potatoes. Y{ ^ 43 Under a tubulated bell- 



glass, Gg, lies the glass ring, Gr. On this is placed the 

 object to be examined, say a potato, K. By means of a cork 

 borer, we make a hole in the potato reaching to its middle, and, 

 after drying the hole with blotting paper, insert in it the 

 cylindrical bulb of a sensitive mercury thermometer, T, graduated 

 to tenths of a degree. It is very desirable for the bell-glass to be 

 provided with two tubulures, the second serving for the reception 

 of an additional thermometer to indicate the temperature of the 



