Holm. — On O bo lari a virginica. 375 
alcohol, the anthocyan loses its colour, while the chlorophyll 
stains the alcohol a rich green. It may be added incidentally 
that it was by this same method that chlorophyll was dis- 
covered in the brownish-coloured Neottia , which for a long 
time was believed to be without chlorophyll. 
The anatomy of the leaf also shows some points of interest. 
The cuticula is rather thick and wrinkled, and the cell-walls 
of the epidermis are strongly undulate, except where the 
strata cover the nerves, at which points they are almost 
straight. Stomata are present on both faces of the leaf, but 
are most numerous on the inferior face. Each stoma is 
surrounded by four or five cells, and is generally parallel with 
the longitudinal axis of the leaf (PI. XIX, Fig. 2). We have 
observed, however, that some of the stomata, especially in 
the neighbourhood of the nerves, showed a somewhat different 
direction, this being due to the stretching of the surrounding 
cells (PI. XIX, Figs. 3 and 4). 
A fact worthy of notice is that the epidermis is but slightly 
coherent with the parts subjacent, this being observed not 
only in the leaves, but also in the stem and, in a very marked 
degree, in the ovary. An examination of the mesophyll 
shows that it contains the chlorophyll equally distributed 
in the various layers. A transverse section (PI. XIX, Fig. 5) 
shows that the mesophyll consists of about seven layers of 
roundish cells; and, strange to say, this tissue does not show 
any differentiation into a palisade or pneumatic tissue. The 
leaf thus becomes isolateral. Although the isolateral leaves 
usually have a palisade-tissue developed on both faces of the 
leaf-blade, a few plants are known in which this tissue is 
entirely absent. This has, for instance, been recorded by 
Heinricher as characteristic of the parasitic genus The shim, 
and it is somewhat surprising to observe a corresponding 
structure in our genus Obolaria. 
The absence of palisade-tissue in The shim has been ex- 
plained by Heinricher as evidently depending upon its 
parasitic nature, but this explanation does not seem tenable 
regarding Obolaria. A careful examination of the chloro- 
