472 
p. L. RICKER 
2. All the leaves on the south side of the trees receiving the greatest 
amount of direct sunlight were densely velvety-pubescent beneath 
with stellate hairs thickly interspersed with long jointed hairs, the 
amount of pubescence gradually decreasing towards the base of the 
branch as the lower leaves received an increasing amount of shade 
from those above, the lowest leaves being practically glabrous. 
3. Leaves on the north side of the tree receiving a minimum 
amount of sunshine showed practically no pubescence except at the 
tips of the youngest unfolding leaves which had a few of the long 
jointed hairs and an occasional stellate hair. Leaves on the east 
and west sides showed intermediate amounts of pubescence. 
4. Leaves on short branches arising from the very base of the tree, 
and in deep shade at all times due to its low dense branching habit, 
were practically glabrous, or with a very few short jointed hairs along 
the midrib beneath. The leaves were also much smaller and of some- 
what different shape and proportions from those on any other parts 
of the tree. 
From the above facts, which have been further verified by an 
examination of plants growing at the Arnold Arboretum and by state- 
ments by Mr. Meyer since his return from China, it can only be con- 
cluded that Castanea mollissima is an extremely variable species and 
that in this species at least pubescence functions primarily as a pro- 
tection to the young growing leaves from excessive transpiration and 
is of no diagnostic value. There has been no opportunity to study 
living material in the fall, but from herbarium material it may be 
concluded that most of the pubescence is lost by that time. It can 
not be argued that leaves on the tree without much pubescence are 
older than those that are densely pubescent and have lost their 
pubescence through age, because the youngest growing leaves, which 
would be the ones to be pubescent if any, are practically glabrous on 
the north side of the tree if protected at all. On the other hand those 
on the south side are densely pubescent in every case, unless growing 
on short branches near the trunk and entirely shaded by leaves on the 
longer branches above. 
These facts serve to show the necessity for a more careful study 
of growing plants and the ease with which a systematist working with 
herbarium material alone may be deceived as to the diagnostic value 
of certain characters. In many instances living wild plants at points 
a few hundred miles apart which are apparently distinct will be found 
