132 J. H. Priestley and Edith E. North 
cases the length of the strips was measured under the microscope by 
the use of a cross wire in the eye-piece and a mechanical stage with 
a micrometer screw traverse. In concentrated sulphuric acid observa¬ 
tion showed that the whole strip increased somewhat in its length. 
As the Casparian strip was thrown into a series of slight regular folds, 
the conclusion is inevitable that the increase in length of the Casparian 
strip as a result of swelling was greater than that of the suberin 
lamellae which enclosed and restrained it. 
In cold concentrated potash the membranes were unaffected but 
on warming a marked undulation was produced and the strips shrink 
in length by a very appreciable amount (5-20 per cent, of their 
original length). This suggests that the inner cellulose lamellae of the 
tertiary Potamogeton endodermis contract most and carry with them 
the Casparian strip, which contracts less and is therefore thrown into 
folds. 
If endodermal strips thus treated are now washed and transferred 
to concentrated sulphuric acid, cellulose walls and the basal substance 
of the suberin lamella both dissolve, the Casparian strip is no longer 
restrained from showing the full swelling effect of the acid and a 
wildly irregular wavy outline is the result (Text-fig. 5). 
(3) The Function of the Secondary Endodermis . 
In many cases the cylinder of an endodermis in the secondary 
stage contains isolated cells, more rarely groups of two or three cells 
vertically above one another (Mylius(i8)), which remain in the 
primary stage. These are the passage cells. Very rarely, as in Spircea 
tdmaria, the numbers of cells remaining in the primary stage is 
exceedingly high; Mylius gives it as 45 per cent., a more usual pro¬ 
portion is one cell in eight or ten. Frequently, especially in the 
secondary and tertiary endodermis of roots, there are no passage 
cells and the secondary endodermis is completely closed until later 
it may be broken by internal growth activities as described by 
Mann ( 17 ) for Draccena fruticosa Koch. If passage cells are present 
they will function in the manner already described for the primary 
endodermal cells. 
If no passage cells are present, the physiological properties of the 
secondary endodermis will depend upon the behaviour of the suberin 
lamella to water and solutes. This subject is very fully discussed by 
Mylius( 18 ) where earlier work is fully dealt with. It is difficult to see 
how Mylius reaches his conclusion that the suberin lamella, whilst 
impeding the entrance of water and solutes, must let them pass; 
but apparently he bases it chiefly upon the fact that he has confirmed 
