996 Groom.- — The Evolution of the Annual Ring and 
ray and annual ring, although it belongs to the section Lepidobalanus ; for 
its broad rays are moderately high, yet slender fusiform structures, which 
are divided near the ends by obliquely vertical lines of fibres into more or 
less separate little rays, and at a greater distance from the ends are undivided 
or deeply penetrated by obliquely vertical thin lines of fibres. Moreover, 
several complex rays of this kind may be almost vertically superposed, 
being separated from one another by only a narrow strand of fibres. 
Thus in this species the rays are more divided than are those of Q. 
lamellosa . 
In Q, incana (PL LXXVI, Fig. 31 a) some of the rays are broad and 
high, and repeatedly divided by many thin vertical, uniseriate or biseriate 
strands of fibres, so that they recall the rays of Q. ( Pasania) densiflora . 
Other broad rays of this species (Fig. 31 b) are slightly thinner and, while 
traversed by less numerous vertical fibres, have oblique, frayed or closed 
bands and uniseriate lines of fibres dividing the ray more or less into super¬ 
posed segments. 
Though there is a general correspondence between the condition of the 
rays and of the annual ring in the different species of Qnercus , and especially 
the American species, there are some deviations from this tendency. These 
deviations may be due to:— 
1. The fact that the broad rays in one and the same species differ in the 
same annual ring, and in the different annual rings. A single broad ray in 
a ring may be represented farther inwards by a complex of separate rays, 
as will be explained later; and possibly in some cases the reverse may be 
true. Thus, in comparing different species of Quercus by this method, 
equal-aged annual rings should be selected for observation. 
2. Difference of climate may in part be indirectly or directly responsible 
for the differences between the rays of the American and Indian species 
of Quercus that are at the same stage as regards duration of leaves or 
structure of annual ring. 
3. The sections Lepidobalanus and so forth may represent not natural 
affinities but collections of relatively non-allied types converging as regards 
structure of fruit. This suggestion, however, is not supported by such 
anatomical evidence as we possess regarding the structure of wood, for 
Abromeit (loc. cit.) showed that in the ‘ White Oaks ’ ( Leucobalanus , a 
division of Lepidobalanus) the walls of the vessels are thin, while in the 
‘ Black Oaks * ( Melanobalanus , or a division of Erythrobalanus) the walls 
of the vessels are thick. 
The question now arises as to the direction in which the evolution of 
the rays has taken place in Quercus and in the Fagales, that is, whether the 
archetype possessed broad rays or only narrow ones. As it is possible that 
evolution can take place in both directions, the answer may be different in 
the case of the genus and the cohort. 
