i8 



except that they were watered only with distilled water. The plants 

 were permitted to grow only as long as they would make good growth 

 in the sand, probably about the time the food supply of the seed was 

 becoming exhausted. Some of the plants of each set were then frozen 

 while others had the sap expressed for osmotic strength (freezing 

 point) determination by the use of a Beckmann apparatus. As a 

 measure of the osmotic strength the term depression will be used in 

 the tables, meaning of course the number of degrees centigrade be- 

 low zero at which, with no supercooling, ice formation begins in the 

 sap. 



Method of Freezing. At first an effort was made to grow the 

 plants in the greenhouse and expose them to outside temperatures 

 to determine the killing temperature. However, this was soon found 

 to be unsatisfactory and the plants were frozen in a chamber sur- 

 rounded by a freezing mixture made of salt and ice. It is evident 

 that the temperature throughout such a chamber would not be uni- 

 form so long as it were falling and great care was necessary to secure 

 as uniform a temperature as possible. The apparatus shown in 

 Fig. 1 was used. In the lower part was an electric fan; the upper 

 part was the chamber in which the plants were frozen. An effort was 

 made to keep the temperature uniform within this chamber by the 

 operation of the electric fan just beneath the hardware cloth shelf 

 on which the plants were frozen. Careful tests showed that the 

 temperature throughout this chamber was always uniform on the 

 same level though sometimes it would vary slightly in different levels. 

 Fearing, however, that this would not always be true, the plants on 

 freezing were always placed, not only on the same level, but at the 

 same distance from the galvanized iron wall of the chamber. In 

 this way it was practically impossible that plants in the freezer would 

 not all be exposed to the same temperature, and consistent results 

 were secured. 



In addition to being sure that the plants were at a uniform tem- 

 perature, it was necessary to control very carefully the rate of fall 

 of the temperature, since rapid falling of temperature very greatly 

 increases the killing. For this reason it was practically impossible 

 to secure results that would be sufficiently accurate so that one freez- 

 ing could be compared with another, except where differences were 

 wide; that is, the plants to be compared must be frozen at the same 

 time. However, when the differences are large it is possible to make 

 a fairly accurate estimate of the relative hardiness of the plants 

 frozen at different times, if great care is taken to duplicate as nearly 

 as possible, the rates of temperature fall. It was found possible to 



