CELLS AND TISSUES 19 



The Periderm. — In both root and stem, there may be derived 

 from the pericycle a phellogen or cork cambium whose derivatives 

 mature as a periderm that serves as a protective tissue following the 

 loss of the epidermis and cortex. Normally, the cells of the 

 phellogen, derived from the outer pericyclic cells, undergo tangen- 

 tial divisions similar to those of a true cambium, differentiating as 

 phellem (cork cells) centrifugally, and phellodermal cells, centripe- 

 tally; but the production of phellem and phelloderm is not neces- 

 sarily reciprocal. Frequently, the number of layers of cork cells 

 cut off greatly exceeds the production of phellodermal layers; 

 and, in some cases, the activity of the phellogen may be unilateral 

 so that only cork cells are produced. The phellodermal cells and 

 the parenchymatous cells of the pericycle centrad to them form a 

 zone which has unfortunately been referred to as a secondary cortex 

 because of its functional and structural similarity to the true cortex. 

 This terminology is misleading, since it implies a cortical rather 

 than a stelar origin of these tissues; and, where such a zone is 

 formed, it is less confusing to denote it as a pericyclic region, or as 

 secondary parenchyma in cases where the pericyclic and phloem 

 tissues cannot be accurately delimited. 



Although the phellogen is commonly derived from pericyclic 

 tissue, this is not always the case. In the young tuber of Solanum 

 an ephemeral phellogen may arise in the epidermis which is soon 

 replaced by one of cortical origin. In herbaceous and young woody 

 stems, the initial phellogen very frequently forms in the subepider- 

 mal layers of cortical parenchyma; while, in older woody axes, 

 the cork cambium develops from secondary phloem parenchyma. 



The Endodermis. — The functional significance of the endodermis 

 in roots is a matter involving much difference of opinion, and 

 several theories have been advanced which have some experi- 

 mental confirmation. The presence of the Casparian strip makes 

 the radial and end walls relatively impervious to the diffusion of 

 water. This suggests that the endodermis may act as a controlling 

 factor in the lateral diffusion of substances between the cortex 

 and stele, since it restricts the passage of material to the protoplast 

 with its semi-permeable and limiting membranes rather than 

 affording free diffusion through the walls. The presence oi passage 

 cells in cases where the endodermis develops to the tertiary stage 

 may be regarded as supporting evidence for this theory. (Fig. 7.) 

 Priestley and North (31) have proposed that the endodermis may 



