Brown : Timber Rot 
11 
from one another (Fig. 26 c). Whether the middle lamella dis- 
solves or splits is difficult to decide, but it would appear that the 
former is the case, in that ruthenium red failed to reveal any trace 
of the middle lamella where the elements had separated. No 
eroding action of the fungus, is to be noted at this stage. The 
infected area is filled with a mass of white cells, in part free from 
one another, or loosely joined at the corners where the middle 
lamella has persisted. 
Before the changes above enumerated have taken place in all 
the outer cells of the infected areas, further alterations usually 
occur in the elements first infected. The cellulose walls now 
undergo digestion (Fig. ii), probably in the usual way through the 
secretion of cytolytic enzymes.^® The vessels, parenchyma cells 
and pith ray cells together with the thin-walled prosenchyma are 
almost entirely dissolved (Fig. 20). The thicker-walled prosen- 
chyma persists the longest, and after the other elements have dis- 
appeared remains as a white fibrous structure partially filling 
and lining the cavities that have arisen. The abundance of these 
fibers depends in a large part upon the cytolytic activities of the 
fungus. The dissolving action may go on in extreme cases until 
the pockets are quite empty of contents, but usually, for reasons 
which cannot be satisfactorily explained, the activities of the fungus 
are inhibited before this condition is reached. 
The conditions met with in the dark-brown peripheral tissue, 
previously described, are quite different from those enumerated 
above. All the cells here have been attacked and partly digested 
(Fig. 7). Many hyphae are to be seen in vertical view closely 
applied to the cell walls and the latter fail to respond to either the 
phloroglucin-HCl or the chlor-zinc-iodide reaction. R. Hartig^ 
noted the same condition in his study of Merulius lacrymans. It 
is possibly explained on the supposition that the fungus has dis- 
solved out but a part of the lignin, and that the portion remaining 
conceals the chlor-zinc-iodide reaction for cellulose. The periph- 
10 Czapeks considers Hgnified walls as made up of a hadromal-cellulose 
ester. The attacking fungus may secrete two or more enzymes, among which 
may be included hadromase and cytase. The first splits off the compound 
ester and removes the hadromal. The cellulose remaining behind is subse- 
quently dissolved by the cytase. 
