1920-21.] 
Interruption of Endodermis. 
very similar appearance to those of root 1 at the same stage in secondary 
growth, i.e. they showed endodermal strands lying between the secondary 
bundles, and also on the extreme outside of the secondary tissue where 
internal cambial activity was very marked (Photo II). 
In both roots the pericyclic cambium, once it was fairly established, 
was most active, whether or no 
it was developed before the 
cortical cambium, since endo- 
dermal cells were always found 
external to the cambium out- 
side the greatest development 
of secondary tissue. There 
appears to be no rule whether 
the cambium shall appear first 
in the pericycle or in the cortex 
just outside the endodermis. 
Both the roots examined show 
that the formation of external 
cambium does not, as Miss 
Spratt suggests, necessarily 
follow after the formation of 
pericyclic cambium, but may precede it. In that case the endodermal ring 
still continues to form a serious barrier to ventilation, though there is 
abundant evidence of intercellular spaces in the secondary tissue. Where 
the cambium originates first from the inside the parenchyma cells pene- 
trating the endodermis 
show air-spaces between 
them (see figs. 3, 4, 5), 
so that the endodermal 
barrier to ventilation is 
broken. But I imagine 
the primary cause of the 
interruption is not to 
allow of greater ventila- 
tion of the inner tissue — though this of course follows — but rather is it an 
adaptation to allow of a greater increase in girth of the root for the forma- 
tion of new vascular tissues, and for a quicker interchange between the outer 
and inner tissues of the enlarging root. If greater ventilation was the prim- 
ary object, then internal cambium would, of necessity, be developed first. 
Professor Scott and Mr Brebner (5) have described a somewhat similar 
