Anatomy of the Cone and Fertile Stem of Eguisetum . 695 
girdled nearly the whole circumference of the stele they find themselves in 
adjacent bundles, and after becoming very closely approximated they die 
out without coming into absolute contact at a height of about i-6 mm. 
above the first appearance of reticulate tracheides. When they die out the 
two strands are separated only by one or two parenchymatous cells. Each 
strand consists of the equivalent of a bundle, but while each may sometimes 
occupy the position of a single bundle it has of course to pass from one 
bundle to another, and frequently occupies a position between two bundles. 
While the masses of nodal wood have been journeying round the stele the 
number of bundles has increased from seven to ten, an increase very similar 
in amount to that found in many normal cones of E. arvense . It is 
interesting to note that this lower annulus, which only faintly recalled a leaf- 
sheath by its lobing on one side, possessed reticulate tracheides apparently 
identical in structure with those of an ordinary node ; the distribution of 
these is of course different from their distribution in an ordinary node ; but 
this is presumably due, in part at least, to the oblique insertion of the 
annulus. In one of the two bundles that first develop reticulate tracheides, 
a bundle situated on the side on which the annulus was most developed, 
there seemed to be indications of a reduced and abortive trace. A few 
phloem cells bend out into the cortex, and endodermal cells with excep- 
tionally well-marked bands of cutinization may be seen cut longitudinally 
a little way out in the cortex. Moreover, one or two tracheides bulge out 
through the opening in the phloem and seem to be passing out into the 
cortex ; unfortunately, two sections, each 10 ^ thick, above the section 
showing these phenomena, were lost in mounting. If one or two tracheides 
did, as seems likely, pass out into the cortex they must have died out 
almost at once, for no further trace of them can be seen in other sections. 
The parenchymatous stelar cells that bend outwards, and those with 
endodermal markings, also die out ere long in the cortex ; but for several 
sections they are distinctly discernible owing to the fact that they are cut 
longitudinally or very obliquely. No further indications of traces were 
observed in any of the other bundles. 
Below the upper normal annulus there were ten strands, and the lowest 
whorl of the cone consisted of ten sporangiophores, so that there was no 
increase in appendicular members ; yet six new meshes were formed above 
this annulus, two of which were closed at or just above the level of the 
lowest fertile whorl. No less than seven out of the ten meshes existing 
below the upper and normal annulus were closed in the region of its 
insertion. Thus the number of meshes closed and initiated in the region of 
the annulus was in the case of Cone F exceptionally high for E. arvense. 
We may add that though annuli do not seem to have been described 
in most of the cones of the fossil Equisetales, Thomas records a possible 
example in Calamostachys (Thomas), and a structure which appears to 
