Physiological Studies in Plant Anatomy 121 
frequently appear thus associated, as in the leaf scar, and that the 
formation of suberin is always dependent upon the presence of 
oxygen. 
The observation of Kroemer ( loc. cit. p. 94) and still more that 
of Rumpf has already associated the appearance of the Casparian 
strip with the differentiation of the phloem. We would suggest that 
this association is causal, and that the Casparian strip is impreg¬ 
nated with substances diffusing from the differentiating vascular 
strands. That these substances are in part of a fatty nature is 
to be expected, such fatty acids forming, according to Hansteen 
Cranner(5,6), an invariable accompaniment of the membranes of 
parenchymatous tissues; Czapek( 7 ) has shown also that they are 
abundant in merismatic protoplasts. They would certainly be lost 
from the contents of the developing xylem elements. When com¬ 
pared with the cases of cutin and suberin formation which are 
better known, notably the formation of suberin at the surface of 
wounds (Priestley and Woffenden( 24 )), the presence of oxidation 
products of fatty acids may be anticipated in the endodermal wall, 
and these will have the general properties of suberogenic or cutino- 
genic acids (Priestley (21)). 
It is very difficult to give experimental evidence of the existence 
of these substances, because micro-chemical data are so indefinite, 
and macro-chemical investigations upon an isolated primary endo- 
dermis have not yet proved practicable. We have tried the effect of 
extracting the vascular strands from the petioles of species of 
Plantago and leaving them exposed to air, or immersed in water for 
some days. After immersion in water for a week the surface of the 
strand is very little altered, but if exposed in saturated air, either in 
the light or dark, within a week it is brown in colour and when tested 
with Sudan glycerine gives indications of the deposit of a substance 
with the staining properties of suberin upon the whole surface of the 
cell membranes at the outside of the strand. 
In view of the physiological evidence to be presented later, 
showing that the radial and transverse walls of endodermal cells 
appear to be only slightly permeable to water and solute and thus 
have properties different from lignified membranes, together with the 
arguments presented above, we conclude provisionally that the 
balance of argument is in favour of the presence in the Casparian 
strip of oxidation or condensation products of fatty acids of the same 
general type as those found in suberin and cutin. 
The experimental work described in this paper was completed by 
August 1921, but some later investigations have provided evidence, 
