CLIMATIC FACTORS : RADIANT ENERGY 119 



on the cell walls crystallizes in the intercellular spaces and con- 

 tinues to crystalize as it is replaced from the vacuole and proto- 

 plasm. Injurious chemical changes, such as the precipitation of 

 proteins, may accompany the desiccation. Some species, however— 

 especially subtropical ones— are often killed before temperatures 

 fall as low as freezing. Temperature injuries cannot always be ex- 

 plained in simple terms. 



It is obvious that there must be seasonal and other adjustments 

 in some plants, which permit their survival as cold weather comes 

 on. It is known, in this connection, that the concentration of the 

 cell sap of most conifers increases in the fall. Gardeners make use 

 of this characteristic, for young plants grown in greenhouses are 

 "hardened" before they are set out and subjected to early spring 

 temperature fluctuations. Such plants are most liable to injury 

 when temperature changes are abrupt and extreme. On the other 

 hand, many arctic and alpine species can grow, flower and fruit 

 during a period when they are subjected almost daily to alternate 

 freezing and thawing. 



Measurement of Temperature.— Accurate standardized glass 

 thermometers are the most useful instruments for field studies. Air 

 temperatures are usually taken in the shade with the thermometer 

 exposed to the wind and away from the influence of one's body. 

 Soil temperatures require a small well of some sort, or, when meas- 

 urements are to be made periodically, a length of pipe may be 

 permanently sunk to the desired depth. If the thermometer is 

 suspended in the pipe by a string, it can be drawn up quickly and 

 read before much change takes place. Soil temperatures at or very 

 near the soil surface are difficult to obtain accurately with an ordi- 

 nary thermometer because of the steep gradients from the surface 

 downward, and upward into the air. The size of the thermometer 

 bulb is sufficient to be affected by rather widely differing tempera- 

 tures even when it is no thicker than 5 mm. Discrepancies have 

 been observed as great as 1 1 ° C. between electrical (thermocouple) 

 and ordinary thermometer readings at the surface. The errors are 

 greatest in full sunlight and on dark soils. 86 It is under these con- 

 ditions that the greatest care must be taken in placing the bulb. 



Continuous temperature records are obtainable with thermo- 

 graphs. These usually consist of an expansion element attached by 



