ioi2 Thompson.—On the Origin of the 
rays of the Dicotyledons have originated by the breaking up of the com¬ 
pound type. 
That the elaborate system of compound rays should have been 
modified seems but reasonable. The influence at work in the development 
of the compound, aggregate, or foliar ray was originally the demand for the 
storage of assimilates descending from the large leaves of the earlier Angio- 
sperms. It was natural that in the warmer climate of Mesozoic time, 
when the leaves probably persisted for a number of seasons, the storage 
devices of the wood should be mainly centred about the leaf-traces. In 
later epochs, however, with the advent of a severer winter season and the 
consequent acquirement of the deciduous habit in connexion with the 
leaves, the organization of the storage system about the leaf-traces as 
permanent centres would no longer be advantageous or desirable. Besides 
being less unwieldy, the system of smaller rays affords equally large 
capacity for storage and a more convenient general relation between 
conducting, supporting, and storage tissues. 
It should also be emphasized that there has been abundant opportunity 
in point of time for the complete evolution of the system of multiseriate 
rays, not only because the compound type is characteristic of the most 
primitive Dicotyledons, but also because multiseriate rays were already in 
existence in the Cretaceous. Stopes and Fujii 1 have described many 
woods of this type from the Cretaceous of Japan. 
The general course of ray development would then be as follows : 
Starting with a primitive uniseriate-rayed condition, the broad rays have 
been built up by a process of compounding primarily about the leaf-trace 
and extending from here throughout the wood. From this condition two 
courses take their origin. On the one hand the continued development of 
the compound ray at the expense of the woody cylinder has resulted in the 
ultimate production of the herbaceous condition. On the other hand the 
forms which remained arborescent replaced the compound rays by a system 
of smaller multiseriate rays with its more uniform distribution of ray 
parenchyma. The typical ray structure of the more recent arborescent 
Dicotyledons accordingly appears to be the multiseriate condition. It is 
obvious that ray structure will supply a valuable criterion in considerations 
of phylogeny and classification. 
Summary. 
i. The production of multiseriate rays by the breaking up of the 
ancestral broad compound type may be observed in many families of the 
Dicotyledons (Ericaceae, Casuarinaceae, Fagaceae, and Betulaceae). 
1 Stopes, M. C., and Fujii, K. ? Studies on the Structure and Affinities of Cretaceous Plants. 
Phil. Trans., Series B, vol. cci. pp. 1-90. 
