Fry . — Some Types of Endolithic Limestone Lichens. 543 
taken not to harm the tissues in any way. If this precaution were not 
taken the microtome knife would suffer badly and accurate sectioning would 
be impossible. Although many hundreds of sections have been prepared, 
remarkably few have been spoilt in this way. 
Sections of all thicknesses between 2 // and 20 ju were examined, but 
for the purpose of this investigation 8 /x was found to be the most suitable. 
Vertical sections of numerous thalli were cut, and from some thalli series 
of tangential sections, each 8 /x thick, were made from the outermost region 
of the cortex to the innermost limits of the rhizoidal zone. 
Heidenhain’s haematoxylin was found to be the most convenient stain — 
either used alone or with Congo red as a counter-stain ; methylene blue 
also proved very useful. Although most of the permanent sections 
have been mounted in Canada balsam, this was not the best medium for 
the examination of such tissues, since the refractive index of the fungal 
walls was so near that of the balsam that their boundaries could only be 
distinguished with difficulty. Some of the clearest preparations have 
been made by mounting the stained sections in colourless glycerine 
jelly. 
By teasing out freshly decalcified material it was possible to study the 
rhizoidal zone, but, owing to the crowding of the surface layers by thalli 
which only occupy small areas, one could not always be certain to which 
thallus the rhizoids under examination belonged, and it was impossible to 
cut sections of material in the freshly decalcified state. There was a further 
disadvantage in using unsectioned material. The hyphae low down in the 
rock did not always belong to the plant occupying the surface at the time, 
for though the previous inhabitant of th,e surface layers might have been 
removed by weathering action, or by the more rapid growth of another 
lichen, its rhizoids remained for some time deep in the rock and apparently 
healthy. 
Classification of Endolithic Lichens. 
A simple classification based on the structure of the thallus is the 
one adopted in this discussion. Although the fruiting organs are not taken 
into consideration in this classification, yet they are described, since they offer 
certain points of interest in connexion with the endolithic habit. 
I. Wholly endolithic -< 
A. Heteromerous. 
(Algal cells — Chlorophyceae.) 
. \ 
B. Homoiomerous . 
(Algal cells — Trentepohliaceae 
or Cyanophyceae). 
II. Partly endolithic 
A. Heteromerous. 
(Algal cells — Chlorophyceae). 
{ ( a ) Cortical zone — loose. 
\b) „ „ fairly compact. 
(c) ,, ,, discontinuous. 
{a) No zonation (only Trentepohlia 
form described). 
I Two parts to thallus : 
(a) Epilithic part — Cortex, gonidial 
layer, and medulla, but no rhizoids. 
( b ) Endolithic part — Accessory goni- 
dial groups plus rhizoidal zone. 
