406 ROYAL SOCIETY OF CANADA 
destroying it, but also give rise to the resultant tissue by a process of 
growth and division. 
That the formation of thyloses, in whatever form it is manifested, 
is due solely to differences in pressure, in the lumens of cells or canals 
which border on each other, is, I believe, a very reasonable assumption, 
proven by the time and manner of thylose development wherever it has 
been observed. This point has been established above for the Ferns, and 
we will now proceed to see if it also holds true for the other forms. 
In the Dicotyledons the thyloses never occur in vessels which are 
in a state of activity, being found only in older vessels. Thus, Stras- 
burger (19, 191), speaking of Robina Pseudacacia states the following: 
“The wider vessels of the three-year-old growth ring were already filled 
with thyloses in the branch investigated.” Further on, in describing 
Ficus, he says (19, 205):—“In the relatively young wood of Ficus 
elastica which was available for investigation, vessels with thyloses were 
observed only at long intervals; on the other hand, thyloses were present 
in all the vessels of a portion of Ficus bengalensis which was obtained 
from an older portion of the stem.” It may be taken as a general rule 
that in the Dicotyledons thyloses arise in the region of the wood which 
is undergoing a change from sap-wood to heart-wood, i.e., in wood which 
is losing the power of conduction. 
The occurrence of the thyloses in this region is just what would be 
expected from the conditions which prevail in the vessels at this time. 
The conductive function is being taken over by the vessels of more recent 
formation, and the sap is thus gradually withdrawn from the older ones. 
The pressure inside the latter is therefore greatly reduced, whilst the 
parenchyma cells which abut on them are in a high state of turgescence. 
The parenchyma cells, possessing a greater turgescence than the adjoir- 
ing vessels, begin to extend, since they are in a state of active growth, 
and the pit being the weakest point in the wall of the vessel, they pro- 
ject into it and push in the closing membrane, finally rupturing it, thus 
gaining access to the lumen. 
The phenomenon is thus identical with that observed in Pteris 
aquilina only in this case there is no free space along the whole conti- 
guous surface of the parenchyma cell, whilst in Pteris the whole side 
of the cell abuts on the cavity. Thus in the former only portions of the 
cell grow into the lumen, whilst in the latter the whole cell takes part 
in the growth and division. 
In the cases where thyloses occur after wounding, or on account of 
disease, essentially the same conditions prevail, for if the various observa- 
tions recorded by authors are scrutinised in this light, it will become 

