GROWTH IN TREES. 41 
Such high temperature, with its attendant conditions of low relative 
humidity and the local maximum of air-flow, would of course heighten 
the rate of water loss of the entire tree. Such water loss would, of 
course, take place from the entire surface, and although it may be 
assumed that the rate per unit area from the trunk would be small, 
it is yet undetermined. 
Water is entering the tree through the outer membranes of the roots 
at all times, in accordance with the balance between the soil mois- 
ture content of the soil and the absorbing capacity of the roots. It 
passes upward by capillarity through the vessels at a rapid rate and 
through the system of inclosed wood-cells or tracheids at a slower rate. 
The mass of these woody cells furnishes a structure, however, upon 
which may be based an explanation of the shrinkage of the trunk. 
Such an explanation must take into account the fact that the rate 
of absorption by the roots shows but little daily variation, due espec- 
ially to the fact that the temperature of the soil may vary but little 
at the depth at which the roots lie. The rate of transpiration, how- 
ever, which may come down to a low minimum during the night fogs 
characteristic of the region, may rise to a high maximum during a 
warm sunny day with a marked air-flow. The water thus lost is 
withdrawn from the wood, either by direct outward exit through the 
bark or upward through the branches and out through the leaves. 
In both cases the withdrawal lessens the supply or balance in the sys- 
tem of tracheids. ; 
The wood of the 1919 layer, when freshly taken from the tree in 
slips less than 2 mm. in thickness, swelled 1 per cent in water and in 
NaOH 0.01N, while the increase was greater in HCl 0.01N, being 
3 per cent of the original thickness. It is thus to be seen that the 
woody tracts through which water passes upward in the tree may have 
a water deficit of 1 part in 100 when swelled in water at a time when 
the actual daily shrinkage amounted to but 1 in 1,700 of the entire 
diameter. It is highly probable, however, that the deficit is unequally 
distributed, being greater in the outermost layers. 
These and other important physical problems are dealt with in 
dendrographic measurements being carried on at the Desert Labora- 
tory, at the Coastal Laboratory, and at other places by collaborators. 
