452 BOTANICAL GAZETTE [DECEMBER 
is to produce pressure on the liquid or gas occupying the pores. If 
liquid alone completely filled the cavities of the wood, any amount 
of thermal expansion would necessarily be accompanied by an equal 
amount of elastic expansion of the wood. Pressure in this case might 
be very great. It should be noted, however, that the pressure recorded 
by a gauge would be less than that developed in the tree without the 
attached gauge, for the sap forced from the tree into the gauge would 
partly relieve the pressure within the tree. It follows that the less the 
amount of sap required to operate a pressure gauge, the higher the 
pressure it will record for a given amount of thermal expansion within 
the tree. 
It probably never happens that the wood of a birch tree becomes 
completely saturated with water. One or two per cent., at least, of 
gas is present in the wood fibers when the wood is densést. A smaller 
amount is present in the vessels. The compressibility of this gas 
lessens the effect of thermal expansion in producing pressure. 
In order to obtain a quantitative statement of the amount of thermal 
expansion, I have made the following estimates. The small white 
birch (11°™ diameter) of exp. 1 has about ro per cent. of its volume 
in small branches and twigs. These must vary in temperature in the 
same way that a blackened-bulb thermometer would, only in a less 
degree—say a maximum daily range of 20°C. The trunk would 
vary less in temperature than the air—say 10°C. The maximum 
daily change of volume of the sap and cell walls of the tree computed 
on this basis would be 270°°, or nearly o.5 per cent. of the volume 
of the tree. Under such conditions the presence of as little as T Pet 
cent. of gas in the vessels would prevent an existing small pressure 
from rising more than about 4o°™. The slight increase of pressure 
due to the greater expansive force of the gas at the higher temperature 
is so small that it may be disregarded. 
We may now consider in detail some of the instances 17 5 
temperature controls pressure by causing volume changes within the 
tree. Fig. 2 is a record of pressure, and of temperature as given by 
a blackened-bulb thermometer. The periods of lower temperature 
were caused by the passing of clouds. Pressure increased during the 
periods of sunshine and diminished during the intervals of shadow- 
From 3:00 to 3:20, while the sun shone bright, the temperature of the 
in which 
