Knox: STEM oF IBERVILLEA SONORAE 335 
(FIGURE 3) many ectocyclic sieve-tubes, but in only one or two in- 
stances has a young commissure appeared. In these cases the 
length of the members of the sieve-tube was that of the diameter of 
the mother parenchyma-cell, the two separated by a large, well- 
formed vertical sieve-plate. The peripheral sieve-tubes are easily 
seen in the very young stem (FIGURES 3 and 4) while the first tracheal 
elements are appearing and the sieve-plates are exceedingly deli- 
cate and distinct. 
The obliteration of the sieve-tubes Fischer treats separately. 
They sometimes. entirely lose their identity, owing to the pressure 
of the surrounding tissues and the torsions of the stem, but some- 
times when the callose breaks down and the contents begin to de- 
generate, there appears in them a slimy substance giving to the 
older stem an appearance which leads him to say: ‘“ One may see 
that we have here another category of cell-derivatives, that our 
commissures and ectocyclic sieve-tubes belong to a hitherto over- 
looked latex-system,”’ and again he says that the presence of these 
ramifying cells in the stem produces the effect of an internal secre- 
tory system. He takes up this point at length in a later paper 
(20) and refers to Hanstein’s theory of the contents of the obliter- 
ated sieve-tubes as ‘‘ Lebensaftgefasse,’’ where Hanstein assumed 
that the contents were active while they were in reality hard and 
gum-like. A section through a stem of /éervillea after its period 
of elongation is completed (FIGURES 9 and 2) shows inside of the 
medullary leptome, peripheral to the outer leptome, and through 
the pericycle and cortex, groups of cells both in structure and 
position identical with the obliterated sieve-elements. They con 
tain a highly refrangible homogeneous substance of a brownish- 
yellow color which makes them very conspicuous. They occur 
ordinarily in pairs, frequently in groups of from three to twelve 
(FIGURES 5 and 6). The contents color orange with hydrochloric 
acid and phloroglucin, while the contents of the active sieve-tubes 
remain lemon-yellow. With Millon’s reagent they turn brick-red. 
Longitudinal sections frequently show the old sieve-plates, and 
the companion-cells seem to have contents identical with those of 
the sieve-tubes. The groups of cells are rarely seen to anastomose 
in young stems, but they frequently runa tangential or radial course, 
and may connect within a short distance inthis manner. In older 
