679 
certain Orders of the Ranales. 
from the lateral bundles of one side of the petiole ; in each case, viz. of 
Paeonia and Calycanthaceae, the midrib-bundle passes into the stem- 
cylinder, while the concentric and inversed collateral bundles respectively 
constitute a cortical system for a less or greater distance down the stem 
(Text-fig. 4). 
Now it is interesting to note that in the majority of plants, when the 
transition from a cylinder to an arc of bundles takes place in the base of the 
petiole, it is the dorsal which may be said to dominate and absorb the ventral 
bundles with which they unite, leaving the dorsal arc as a result of the pro- 
cess. But in the Calycanthaceae precisely the reverse happens : for the 
ventral bundles dominate and absorb, so to speak , the dorsal, with the result 
that a ventral and, therefore, necessarily inverted , system of bundles is left 
over which persist right through the entire internode ; the large size of each 
inverted bundle is due to the fact that it represents the fusion of two ventral 
bundles. It may be objected that no ventral bundles which unite with the 
dorsal actually exist in the leaf-base, as in other orders. This is true : but 
what I have suggested above represents what I believe has occurred in the 
phylogenetic sequence of events, the probable successive steps of which may 
be indicated as follows : — 
(1) The ventral bundles of the petiole-cylinder pass across and fuse 
with the dorsal ones, a dorsal arc resulting ; 
(2) a half-way stage, represented to-day by Paeonia , Berberidopsis , &c., 
in which both ventral and dorsal bundles are equally strong and both 
refuse, as it were, to effect a complete union ; 
(3) the dorsal bundles, with exception of the midrib, pass across and 
unite with the ventral, a ventral arc, complete only in its lateral portion, 
resulting ; 
(4) the ventral bundles are pushed further and further downwards until 
they disappear from the leaf and occur only in the cortex of the stem, as in 
Calycanthaceae. 
Hence, it appears to me to be not strictly, but rather only metaphoric- 
ally, true to say that the inverted cortical bundles are leaf-traces in the 
sense that they emanate directly from the leaves. My own observations 
certainly show that, as regards the ontogenetic history, four lateral bundles 
from the petiole-arc pass into the cortex of the stem and unite with strands 
which are already in existence there as, in a sense, cauline strands. These 
latter are far too large in size to belong to the lateral bundles of the petiole. 
My conclusion is, that they belong in present actuality, though not phylo- 
genetically nor essentially, to the stem and not to the leaf. Phylogenetically, 
they are leaf-traces. This is in agreement with the conclusions of Van 
Tieghem and Herail ; the older authors like De Bary are, I am convinced, 
in error in regarding these bundles as leaf-traces. 1 
1 Herail says in effect that the inverted cortical bundles of Calycanthaceae are not derived from 
