i75 
Vessels of the Onion. 
than the thin side wall, stained like callus, though somewhat 
faintly, for about a third of its length in its central portion, 
while just at the end of the stained part a very small more 
brightly staining blunt peg projected into the cell on one 
side (Fig. 7). 
This and similar cases are of interest as regards the 
question whether callus is formed from the cell wall or secreted 
directly by the protoplasm. With regard to the participation 
of the wall in the formation of callus the following differences 
were noticed : — 
1. The wall itself is not changed, but the callus is super- 
posed sometimes only on one (Fig. 3, a ; and 2, a), sometimes 
correspondingly on both sides (Fig. 3, c, d\ and 2, 3). 
2. The wall gives the callus-reaction, but more faintly than 
the superposed formation, and it can by virtue of this be 
traced through the callus masses where these correspond in 
position on both sides of the plate (Fig/3, a, b , d). 
3. The wall cannot be distinguished at all ; the staining is 
of a uniform tint throughout (Fig. 3, b , c). 
The first case is more frequent where there is only a small 
callus-formation ; either the second or the third usually ob- 
tains where the callus is more extensive, but sometimes where 
only very partial. With very extensive callus-formations, the 
side walls sometimes stain for a little distance beyond the 
transverse septum, the colour of the stain becoming fainter 
and fainter and gradually passing into the ordinary non- 
staining cellulose (Fig. 3, b, lower plate). These facts point 
to a change induced in the cellulose of the wall by which it 
becomes more (as in 3) or less (as in 2) converted into callus, 
as indicated by more or less deep staining, while pure callus 
is deposited on this by the protoplasm in the same way as 
thickening on a wall. 
Function. As stated above, these laticiferous cells contain 
none of the ordinary food-stuffs in any perceptible quantity, 
nor have they any connection with the assimilating paren- 
chyma on the one hand, or with the vascular bundles on the 
other. There seems to be no direct relation between presence 
