THE LONGEVITY OF TREES. 85 
section, from the circumference towards the centre, where the 
space in which the layers could not be clearly made out was 
estimated to have comprised three hundred more. If the in- 
jured portion was not overestimated, the tree must have been 
a thousand years old. We have now before us a section of a 
fine trunk of the American Cypress ( Zaxodium distichum), 
upon the radius of which, twenty-seven inches in length, six 
hundred and seventy annual layers may be distinctly counted. 
The wood of this tree is so durable, that probably the age of 
trunks of more than twice that size might be ascertained by 
direct inspection. 
When such a section cannot be obtained, we are obliged to 
resort to other and less direct evidence, affording only ap- 
proximate, or more or less probable, conclusions. Sometimes 
lateral incisions, not endangering the life of the tree, furnish 
the means of inspecting and measuring a considerable number 
of the outer layers, and of computing the age of the trunk from 
its diameter and actual rate of growth. But as young trees 
grow much more rapidly than old ones, we should greatly ex- 
agegerate the age of a large trunk, if we deduced its rate of 
growth from the outer layers alone. We must therefore 
ascertain, by repeated observations, the average thickness of 
the layers of young trees of the same species; and by the 
judicious combination of both these data, a highly probable 
estimate may often be formed. 
When unable to inspect any portion of the annual layers of 
remarkable old trees, we may occasionally obtain other indi- 
cations upon which some reliance may be placed ; such as the 
amount of increase in circumference between stated intervals ; 
but as, on the one hand, we can never depend upon the entire 
accuracy of two measurements made at widely distant periods, 
while, on the other, the growth of a small number of years, 
however carefully ascertained, would be an unsafe criterion, 
this method can seldom be employed with much confidence. 
A more common mode is to employ the average rate of growth 
of the oldest trees of which complete sections have been ex- 
amined, for the approximate determination of the age of re- 
markably large trunks of the same species, where the size 
