in the Bronzes and their Brown Rust. 309 
_ Columns 5-7 were introduced in order to give the data 
for obtaining the approximate calculations in columns 8 and 
9, viz. calculations of the mean area of chlorophyll-tissue 
and epidermis divested of the vascular tissue and alone 
attackable by the Fungus. I tried several methods of more 
direct measurement of this tissue before deciding that the 
irregular shape of the leaf defeated all attempts to obtain 
such. But it appeared so necessary to ascertain, however 
approximately, whether any relation exists between the 
volume of attackable tissue per square unit of leaf-area, and 
the susceptibility or immunity of the species, that I hit upon 
the following device. 
Having subtracted the width of the vascular bundles 
(column 5) from the total width of the leaf (column 4) as 
measured on sections, I took the mean depth of the leaf- 
section from columns 6 and 7, and then treated the whole as 
a rectangle. The area of this gives us the result shown in 
column 8, reduced to a rectangle with its longer side equal 
to one millimetre. This sectional area multiplied by 1,000 
gives us the volume of tissue per 1 sq. mm. of leaf surface. 
It must be admitted that these areas and volumes cannot 
be exact. The ideal object of attainment would be the 
volume of the cells and inter-cellular spaces, numbers of 
chlorophyll-grains and their sizes, and so on, but I found 
it hopeless to expect any success in carrying out such estima- 
tions as were attempted in this connexion. 
One point that has to be attended to is to allow for the 
inter-osculations of vascular bundles by lateral bundles. It 
is quite a mistake to suppose that the * veins * of grasses are 
all parallel ; they are netted, and the only device I could 
think of for allowing for the lateral connexions was to deduct 
the tissues immediately above and below the much-predom- 
inating main ribs. In any case, this is the least attacked of 
all the non-vascular tissue. - 
Perhaps all I can claim for this part of the subject is to 
have suggested a method which can be improved upon in 
future measurements. If the latter were made on a sufficiently 
