134 J- EL Priestley and Edith E. North 
be argued that its obstructed passage was due to its own precipi¬ 
tating qualities. Similar experiments were therefore carried out with 
other salts, using large Monocotyledonous roots having a tertiary 
stage endodermis without passage cells, also the rhizome and fronds 
of various Polypodiacese where the secondary endodermis was un¬ 
broken by passage cells. In some of these experiments lead, iron 
and copper salts in strong concentration were driven up the vascular 
strands under pressure, and subsequent micro-chemical tests showed 
that the cells within the endodermis were completely penetrated 
but none of the toxic salts were present in the cells of the endo- 
dermal ring or outside it. In other experiments the toxic salts were 
allowed to diffuse in through the exposed cortex, and the endodermis 
then proved a barrier to their inward diffusion. 
It is difficult to perform critical experiments upon the permeability 
of the suberin lamella to water, but the resistance of the suberin 
lamellse of the periderm to evaporation is almost an axiom among 
botanists. Mylius summarises what evidence there is to show the 
impenetrability of the dead suberin lamella in the periderm to gases, 
a property which, in view of the gas enclosed within the dead cells, 
would give the cork layer good insulating properties. Haberlandt(iO) 
(loc. cit. p. 137) gives experimental evidence for the general reduction 
in evaporation produced by the cork layer as a whole. As differences 
have definitely been established between the suberin lamella in the 
endodermis and this lamella in the periderm, it is not possible to 
take over these experimental results bodily, but a general similarity 
of behaviour of the two lamellae may be expected. This expectation 
seems to be borne out by the result of the following experiment. 
Some lengths of the thick prop roots of a Palm were taken, and the 
thick impermeable outer layer removed. These roots possess an 
endodermis in the tertiary stage without passage cells. The lengths 
of roots were divided into two groups, in one group the cut ends 
were sealed in paraffin wax, whilst the ends of the other group were 
left unsealed. Both groups were weighed, then placed in a desiccator 
over calcium chloride and weighed again at intervals. The unwaxed 
pieces lost weight most rapidly and in a few days reached a constant 
weight, the waxed pieces still lost weight slowly for another week 
and then w r ere cut in half transversely and left to reach a constant 
weight also. The exposed end of the waxed pieces were then sealed 
up and all the pieces transferred to water and left for nearly two 
weeks. At the end of this time the unwaxed pieces were within 
o*2 grm. of their original weight, whilst the waxed pieces were still 
1 grm. (about 10 per cent.) short of their original weight. This 
