IOIO 
Thompson.—On the Origin of the 
a vertical series of lower rays. The beginning of this process is shown 
in PI. LXXVIII, Fig. 20, which is a photograph at the same magnification 
of a section taken further from the pith of the same branch. Still further out 
the separating strands increase in size, with the result that the low rays or 
segments of an original high ray are gradually pushed apart. This con¬ 
tinues until the condition found in the adult (Fig. 21) is attained. Here 
the rays are scattered through the wood and give little suggestion that 
they have originated from the oblique dissection of vertically elongated 
compound rays. Owing to the obliquity of the dissecting fibres the pro¬ 
cess cannot usually be well observed in transverse sections. Occasionally) 
however, the fibres are sufficiently vertical to make the process visible 
in favourable sections. This is especially true in the European Beech, 
Fagus sylvatica. In Fig. 22, taken from a section of the latter species, 
the broad ray in the centre of the field is undergoing the oblique division. 
Even in certain species of the oak itself, in which the broad rays reach 
their maximum development, the beginning of their dissection may be 
seen. In Quercus bicolor , for example, in the later formed wood of very 
old trees the rays begin to undergo oblique division. 
In Platanus occidentalis the conditions are almost identical with those 
described for Fagus ferruginea. 
If the smaller multiseriate rays have originated as suggested, it is 
to be expected that typical multiseriate-rayed forms would exhibit some 
evidence of their derivation from a broad-rayed ancestry. Such evidence 
has been found in many forms. As a first example we may take the case 
of Carpimts japonica. This species is characterized by the possession 
of multiseriate rays. Yet occasionally one may find in the seedling the 
broad oak type of ray. Fig. 23 shows such a ray at the bottom of the 
field dividing to form the two smaller rays at the top. Obviously we 
have in this primitive region a recapitulation of the process by which 
multiseriate rays have originated. 
Ostrya virginiana , the hornbeam, which is likewise characterized by 
the possession of multiseriate rays, furnishes a second example. Fig. 24, 
from the root of a seedling, shows the retention in this species also of the 
broad type and its division into the smaller type. Moreover, wounded 
seedlings of Ostrya occasionally exhibit broad rays in the wood formed 
immediately after injury. This phenomenon is to be regarded as a trau¬ 
matic reversion and, together with the observation just recorded, is inter¬ 
preted to mean that the multiseriate-rayed Ostrya is descended from 
forms which possessed broad rays. 
The conditions in the genus Betula , which has been investigated by 
Professor Jeffrey, lead to exactly the same conclusion. Species such as 
B. populifolia (Marsh) possess compound rays of the so-called false type 
throughout the ordinary wood. Others, such as B. alba (L.), possess them 
