of the Endosperm during Germination. 
451 
Examination of the cell walls about the region where this 
change is in progress shews that the action of the ferment on the 
walls does not proceed in a centrifugal manner, as was seen to 
be the case in Tamus ; for instead of the ferment originating in 
the embryo and proceeding outwards to the endosperm, it appears 
to originate in the endosperm and proceed inwards towards the 
embryo, and so work in a centripetal manner. This method of 
attack, which can easily be followed with the help of suitable 
reagents, such as Congo red or picric aniline blue, gives a 
very peculiar and characteristic appearance to sections of the 
endosperm of germinating seeds (Fig. 13), for the walls which are 
attacked shew stained mucilaginous areas on their sides towards 
the periphery, i.e. away from the centre of the seed, whilst the 
sides towards the centre retain their sharp contours and shew little 
sign of disorganisation (Figs. 14 and 15). 
The relation between the progress of the ferment action and 
distribution of the “ connecting threads ” is not always very clear, 
for though in some cases a boring out along certain of the threads 
can be observed, yet in many others the dissolution of the walls 
appears to take place without any obvious connection between them. 
Cell walls on which the ferment has just commenced its attack 
frequently shew a jagged edge, owing to the ferment forming little 
V-shaped disorganized areas, which are at first separated from each 
other by thin slips of the unaltered cell wall (Fig. 14), and in such 
cases the action appears to take place along the canals of certain 
of the connecting threads, but when the disorganization is more 
advanced the mucilaginous portions of the wall are usually found 
to be bounded by rounded surfaces of the unaltered cell wall, and 
it is then difficult to make out that the ferment action proceeds by 
means of the threads (Fig. 15). The disorganization of the wall 
gradually extends until only a few islands of cells situated in 
a mucilaginous matrix remain, shewing the original composition of 
the wall, and these finally disappear, leaving an uniform wall of 
hydrated cellulose in which groups of connecting threads can still 
be clearly seen (Fig. 14). 
When germination has proceeded for some time sections of the 
endosperm shew that the centripetal action of the ferment has 
commenced in cell walls at some considerable distance from the 
embryonic cleft, especially in cells under the testa (Figs. 13 and 14). 
The progress of the ferment in these distant walls is essentially the 
same as that just described, and sections treated with methylene 
blue shew that there is a very large unstained area around the 
embryonic cleft. With the vigorous growth of the embryo the 
cotyledons increase in size and continue to encroach upon the 
endosperm until the reserve materials stored in the cells and cell 
walls are completely absorbed, and, as soon as this has been 
32 
VOL. XI. PT. VI. 
