19G 
PROFESSOR W. C. WILLIAMSOR" ON THE ORGANISATION 
medullary cells coiild be discovered, I bad little hesitation in concluding that the small 
twigs devoid of medulla and the Jarger ones, in which such a medulla was very con¬ 
spicuous, belonged to the same plant. This question of the development of a medulla 
in a manner so different from what is seen amongst living Exogens has hmg required 
to have more light thrown upon it; and I propose, in the present Memoir, to record 
some additional observations that I have made on the subject. But before doing so I 
would call attention to the existence of two distinct modes of ramification amongst 
these Carboniferous Lycopodiaceous plants. In one group, illustrated in Memoir 
Part III., Plate 43, figs. 19 and 20, and also in the stems of the Arran Lepiclodendron, 
the vascular cylinder (etui medullaire) presents a dichotomous ramification, in which 
the cylinder divides into two virtually equal horseshoe-like halves. But in other 
instances only a small vascular segnient separates from the cylinder. In my Memoir, 
Part II., p. 224, I showed that in Halonia segments only of the medullary vascular 
ling were detached to supply vascular bundles to the tubercles so characteristic of 
the genus, and which I showed {loc. cit., pp. 224-5) were merely branches that had 
undergone an arrested development at an early stage of them growth. In the 
Memoir, Part XII., Plate 32, figs. 22, 23, 24, and 25, I showed that a similar mode of 
ramification existed in a Halonial {i.e., fruiting) branch of the Arran Lepidodendron. 
In the classic species of Lepidodendron, L. Harcourtii, I have not as yet succeeded 
in discovering any example of the first of the above modes of dichotomisation. But 
my young friend, Mr. Lomax, of Badcliffe, brought me a fine branch of this plant in 
which the second tyjDe was conspicuous. This specimen gave us eleven transverse 
sections, which successively showed the progress of a bundle from its first separation 
from the medullary vascular cylinder to its existence as the vascular axis of a 
branch—Plate 5, fig. I, represents a portion of the medullary vascular cylinder, as 
seen in the lowermost of tlie eleven sections. At a we have part of the large paren¬ 
chymatous medulla. At h) h we see parts of the continuous ring of the medullary 
vascular cylinder. At h' we discover a segment of that cylinder becoming detached 
from the remainder at the points 6", h”. In fig. 2 we have the portions a and h, h of 
fig. 1, hut the segment h' is now completely separated, as represented in fig. 3. The 
vascular portion is hut little altered at h', h', whilst a small portion of the medullary 
parenchyma, a, of fig. 1, coheres to the detached vessels at fig. 3, a. In fig. 5 we 
find the segment changing its form. Its two points b', b' are curving inwards towards 
one another, l;)ut traces of the medullary cells are still seen at a. In fig. 4 the con¬ 
vergence of the two points b', b' has advanced so far that they now virtually touch 
one another, wdiilst scarcely any traces remain of the medullary cells at a. They 
seem to have undergone absorption. In fig. 6, which now represents the transverse 
section of the vascular bundle as seen in a separating branch, the coalescence of the 
points b', b' of the previous figures is complete. The bundle has now attained the 
form of a symmetrical cylinder, wholly composed, apparently, of scalariform vessels. 
Having thus become t-hc ascuiar bundle of an ordinary branch of a stem of which 
