JUNE 7, 1901.] 
In no case examined, however, does a 
bundle end ina gland, as held by Figdor 
(.¢.). The relations of the glandular and 
vascular tissues are none the less intimate 
and striking, as we shall see in coming to 
the second point. 
Fic. 5. Diagram of a transverse section through 
a nectary. ep, epidermis ; ¢c and e, as in Fig. 4. 
In examining a transverse section we see 
the glandular tissue extends towards and 
comes partially to surround the vascular 
bundles adjacent (Fig. 5), and, further, that 
the pericycle of these bundles, in the regions 
contingent upon the gland, consists of three 
to four layers of enlarged cells, and not, as 
is to be found elsewhere, of asingle layer of 
cells (Fig. 6,a@ and b). These enlarged 
pericycle cells are abundantly supplied 
with cytoplasm of fine, tenuous structure 
with large and often irregularly shaped 
nuclei. The cytoplasm is little vacuolated. 
Fic. 6. a, transverse section through a portion 
of the pericycle in contact with the glandular tissue. 
b, a similar portion in a bundle adjoining the ground 
tissue. ¢, cribral cells ; p, pericycle ; e, endodermis ; 
g, ground parenchyma. 
The pericycle cells found elsewhere in the 
petiole, where the glands are active, are 
supplied with ascant amount of cytoplasm. 
SCIENCE. 
889 
The endodermis, * which is usually regular 
and well marked, is, in the vicinity of the 
nectary, quite irregular and often difficult 
to recognize, and its cells have very much 
the character of the adjoining gland cells 
in shape and content. We may thus 
regard these cells, namely, those of the 
endodermis and pericycle, where the bun- 
dles and nectary touch, as a part of the 
gland. . Whether we may assign a different 
function or the same to these cells may not 
be answered. Bonnier (J. ¢.) has indicated 
that, besides the two sugars above named, 
au invertin is to be found in the gland. 
The suggestion occurs to us that this sub- 
stance originates in the pericycle. Never- 
theless, the greater development of the peri- 
eyele as already indicated can perhaps be 
entirely accounted for by the activity of its 
cells in passing the soluble carbohydrates 
from the moving sap into the gland. 
After the activity of the gland ceases, the 
thickening of the walls of the endodermis 
commences, in all the bundles lying in the 
petiole near the fork formed by the pinna, 
and the process extends to the surrounding 
parenchyma. In this way the strength of 
the fork is very materially heightened. We 
notice also, with Figdor, a thickening of 
the walls of the gland cells, a process 
which takes place in all the chlorophyll- 
bearing cells beneath the stomatal bands. 
In offering a teleological explanation of 
the organs above described, F. Muller 
pointed out that in Brazil the fern is visited 
by an ant (Cremogaster sp.) of which a leaf- 
 eutting species,an Oecodoma, standsin dread. 
To this Francis Darwin } answers that the 
plant has few natural enemies—meaning, 
presumably, in England, though this is 
equally true, so far as observation goes, 
in North America. Francis Darwin further 
suggests, in the view of the possible weak- 
ness of the above explanation, that the 
* Derived from the phloeoterma of Strasburger. 
+ Nature, 16: 100. 
