338 -© The American Naturalist. | [April, 
there is formed a large knob, which is the ‘‘ knee’ under discussion. 
This knee, when fully developed, is generally hollow, comparatively 
soft, gnarled, and very difficult to rupture, so that it has the quality of 
a spring that becomes more rigid as it is extended or compressed out of 
its normal shape. When in a hurricane the great tree rocks back and 
forth on its base, and with its immense leverage pulls upon this odd- 
shaped wooden anchor, instead of straightening out in the soft mater- . 
jal, as an ordinary root might, thus allowing the tree to lean over and 
add its weight to the destructive force of the storm, it grips the sand 
as the bower-anchor would do, and resists every motion. The elasticity 
at the point of junction allows one after another of the perpendicular 
flukes attached to the same shank to come into effective action, so that 
sea 





A Yep 
INN 
HYPOTHETICAL! 
HS 
Bae / 
Y Wy y l / j 
YY YY) 
From ‘‘ Garden and Forest.” Copyright, 1890, by the “ Garden and Forest Pub. Co.” 
Ss 

before being drawn from the sand or ruptured the combined flukes 
present an enormousjresistance. 
The above drawing I have made for the purpose of simplifying the 
discussion. It shows a hypothetical Cypress with two roots of the same 
length and diameter—one with knees, the other without them. The 
superior strength of the stiffened root would seem sufficiently evident ; 
but, with the view of obtaining the judgment of a mind thoroughly 
trained in questions of this nature, I submitted the drawing to my 
friend, Charles Macdonald, late Director of the American Society of 
Civil Engineers, whose eye has been accustomed to estimating the 
value of strains in structures by an active experience of twenty-five 

