[KIRscH] CERTAIN STRUCTURES IN THE PTERIDOPHYTES 403 
That this is not merely a supposition, but a probable fact is con- 
firmed by the following observations of Pfeffer on the effect of mechan- 
ical pressure on growth (12, 28) :—‘ Mechanical resistances exercise a 
direct retarding action upon growth, and cause it to cease when they are 
sufficiently powerful, but at the same time they may act as physiological 
stimuli. The growth of the cell-wall may continue even though external 
growth is impossible, and as soon as the mechanical resistance is removed 
growth is resumed in those zones which remain meristematic.” This 
certainly applies to the phenomenon under discussion, for in this 
case the rapid movement of the transpiration stream not only offers 
mechanical resistance to the extension of the parenchyma cells, but also 
acts as an irritation and in this manner the cells are kept in a state of 
great internal activity. 
The cells bordering the canal are constantly receiving large supplies 
of nutritive material and are in a high state of turgidity. During the 
time that the sap ascends only by way of the intercellular canal, how- 
ever, the pressure exerted by it on the wood parenchyma ceils greatly 
exceeds the pressure arising from their highly turgid condition and pre- 
vents them from extending into the cavity. When this force is dimin- 
ished, their turgidity exceeds the pressure exerted against them, and 
they begin to grow into the canal, and, since a large quantity of nutri- 
tive substances is available, they divide rapidly. 
This explains why the thylose formation is initiated and gves on 
most energetically just as soon as the xylem begins to assume its adult 
characteristics, for the lumen of the canal is then available for this pur- 
pose. The canal therefore not only functions as a vessel, but, as in the 
case of vessels, becomes blocked with thyloses when it begins to lose its 
function. It will be shown below that all thylose formation, whether 
normal or abnormal, is probably due to the same cause. 
In regions of slow growth, where the canal is not very large in the 
first place, and where the xylem slowly assumes its functions, the 
change in the pressure is very gradual, and cannot be very great, so that 
thyloses rarely make their appearance at an early stage; the slow for- 
mation or entire absence of the thyloses in these regions would also be 
due, in part at least, to the less active condition of the parenchyma cells 
which bound the canal, since all the tissues here are comparatively in- 
active. 
That the observations in the Equisetaceæ and the Monocotyledons 
are capable of being interpreted by means of this explanation, is, I be- 
lieve, evident from the following: In both these groups the xylem plate 
in each bundle is much less extensive than in the bundles of those ferns 
where the phenomenon of thylose formation was most in evidence. Thus 
