Points in Grading 81 
of the requirements of tree life, for if the cover is of sufficient 
depth, say two feet or more, and especially if of compact 
soil, it will usually quickly and surely kill the tree. 
No less ignorance is exhibited by those who, when such 
filling is to be done, leave an unfilled depression or space 
directly around the bole, perhaps walling it up with brick 
and covering with wire netting or filling with rubble. In 
doing this the notion is to admit rain water, and perhaps 
also air, to the roots. But, while undoubtedly the admis- 
sion of water and air is facilitated by this device, the need 
of water and air is much less near the bole where only con- 
ducting roots are located, than on or towards the ambitus 
of the crown, where the feeding rootlets lie and the respira- 
tion is most actively needed. Filling around the bole is, 
to be sure, undesirable as it not only clogs respiration in 
that part of the bole, but, with some species, induces the 
formation of suckers, which weaken the tree. Hence the 
precaution of keeping the base of the bole free is a good 
one, but it does not compensate for the filling over the 
feeding rootlets. 
There are three ways of meeting the trouble, if there is 
a strong desire to save the trees instead of removing anc 
replacing them by others. Where practicable, the filling 
should be done so as to leave depressions with gentle slopes 
towards the feeding area cf the root system; or better still, 
the filling should be done gradually, taking if possible 
several years, and after first thoroughly stirring and fertiliz- 
ing the ground around the roots, so that the tree may make 
new feeding roots towards the surface. 
Or, where deep fillings have to be made and valuable 
trees are to be saved, the following device may be employed. 
After the ground has been stirred and fertilized as suggested, 
an air chamber may be established by covering the ground 
