Medullary Rays of Quercus. 997 
There are three possibilities, namely, that in addition to the narrow 
uniseriate rays the archetype had:— 
1. Rays broad and high from their commencement (including broad 
primary rays). 
2. Exclusively narrow rays in the inner annual rings, and broad rays 
farther from the centre, 
3. No multiseriate rays. 
In either of the first two cases broad rays have given way to narrow 
ones in some types of Quercus . 
The condition of the rays in the two families of the Fagales gives no 
reliable clue to the primitive state, for in each family there are broad-rayed 
( Quercus , Fagus , Alnus) as well as narrow-rayed ( Castanea , Castanopsis , 
Alnus) types. It is thus evident that whichever direction evolution has 
followed, the change has taken place independently in both families, and in 
both families transitional rays occur 7 for instance, in Quercus , Alnus , Carpinus , 
and Corylus . In Castanea vesca the rays are shallow and uniseriate. In 
Fagus sylvatica y in addition to uniseriate rays, there are more or less 
undivided or divided multiseriate, as well as grouped or isolated small rays 
varying from uniseriate linear to pluriseriate fusiform (2-15 cells thick). In 
Alnus glutinosa, besides the numerous uniseriate rays, there are ‘ false rays ’ 
composed of densely aggregated largely uniseriate, but also biseriate rays. 
In Carpinus Betulus the individual rays are thicker than in Alnus gluti- 
nosa y and vary from uniseriate to thin fusiform rays, which often have uni¬ 
seriate ends and a median fusiform portion 2-3 cells thick: there are 
also rays that are 2-3 cells thick near their ends, but uniseriate in the 
middle, and thus represent two superposed fused rays (or one that has not 
yet divided): the { false rays ’ are aggregates of such rays. In Corylus AveU 
lana both the individual and ‘ false rays ’ are thicker in turn than in Carpinus 
Betulus y the biseriate and triseriate being more numerous relatively and 
being supplemented by quinqueseriate rays. In Betula alba biseriate and 
triseriate fusiform rays are abundant. 
If we assume, as seems most probable, that the Fagales represents 
a reduction-series as regards floral characters, and that Castanea is in this 
respect the most primitive of the cohort, it would seem not unlikely that 
the wood of Castanea should also represent the primitive condition in the 
cohort. Now, Castanea has narrow rays : in Castanopsis indica the abun¬ 
dant uniseriate rays are supplemented by occasional groups of small 
fusiform (up to 5-seriate) rays, which are ranged in tall vertical bands that 
agree structurally with the c vertically divided broad rays ’ of (Pasania and) 
Quercus y and the wood of the latter shows many stages that may be 
interpreted as denoting a compounding of fine rays to form broad ones. 
(En passant , it may be remarked that if in the ancestral Quercus the rays 
were all narrow, then if we are to judge by modern types this archetype 
