KNOWLEDGE OF MONOCOTYLEDONOUS SAPROPHYTES. 185 
sisting of one layer of cells. Mycorhizal hyphæ may be seen out- 
side the roots and penetrating and traversing many of these cells. 
(2) Exodermis (Pl. V. fig. 5).—This forms a remarkably distinct 
and regular layer of cells. Most of its cells are elongated 
longitudinally, and their radial diameters are longer than their 
tangential; they have very thick, stratified, cuticularized walls, 
with middle lamellx, which are lignified excepting towards the 
subjacent cortical cells. The lateral and internal walls have 
‘simple pits; the external walls are not pitted. Scattered amongst 
these cells are the smaller passage-cells, with thin lignified walls 
and dense protoplasmic contents. Each passage-cell is shaped 
like a truncate cone with the broad base inwards. Its lateral walls 
have numerous simple pits, placing it in communication with the 
adjoining exodermal cells. To render the inflow of liquids still 
more easy, the inner wall of the passage-cell is frequently pitted. 
The majority of the mycorhizal hyphe pass in vid the passage- 
cells, but a few pierce the other exodermal cells. 
Within the exodermis there succeed about eight layers of 
-cortical parenchyma, the cells of which have thin, pitted, cellulose- 
walls. Of these about four layers form an 
(3) Exocortex, which is made of cells which are more elongated 
and narrower than those comprising the inner portion of the 
cortex. Many of these cells have coiled mycorhizal hyphs, 
and amongst them occur scattered raphide-mucilage cells. In my 
specimen, which was in blossom, no starch occurred in these cells. 
(4) The inner portion of the cortex is constituted of broader 
and shorter cells, excepting those immediately outside the 
passage-cells of the endodermis, which are narrow. In mature 
parts of the roots of my specimen neither starch nor mycelial 
masses occurred. In the youngest parts of the roots mycelial 
masses were to be seen in all stages of formation. The change 
in the nuclei of the cells in which the masses were commencing 
was striking. The nuclei evidently increase in size, and become 
packed with relatively very large, deeply staining granules. In 
these young parts of the root mycelial masses oceurred in all 
the layers of the inner portion of the cortex, including the 
layer immediately outside the endodermis. Thus there is no 
distinction into a mediocortex and an endocorter. It is to be 
noted that in the mature parts of the roots of this plant not a 
grain of starch was visible, nor was there a trace of mycelial 
masses. This plant was just at that stage during which it would 
