56 
Hoar.—A Comparison of the Stem 
The family Cornaceae is composed, according to Engler and Gilg’s 
classification, of three sub-families, namely, the Mastixioideae, the Curtisioi- 
deae, and the Cornoideae. From this family they separate the sub-families 
Nyssoideae, Davidioideae, and Alangioideae, placing them under the order 
Myrtiflorae. As I shall attempt to show later, the evidence for such 
a separation is lacking when the anatomy is taken as the criterion. 
Of the various genera of the above sub-families, I have been able 
to secure examples of Cornus, Nyssa, Davidia , Griselinia, Aucuba , Corokia, 
Mastixia , and Helwingia . In the case of Cornus , Nyssa, Griselinia , and 
Aucuba , I have had specimens of the roots as well as of the stems. 
Regarding the general anatomical characteristics of the Cornaceae we 
are indebted chiefly to the work of Sertorius (3). It would seem best, perhaps, 
to insert here some of the more important anatomical characters which he 
found in his study of the stem axis. The cork arises usually sub-epidermally, 
though Moller and Weiss state that the phellogen in Aucuba and in some 
species of Mastixia arises from the epidermis. Collenchyma appears in the 
cortex in nearly all cases, though none has been reported for Mastixia. 
Cortical bundles appear only in the case of Mastixia , and are due there to 
the leaf-trace which passes down in the cortex for a short distance before 
entering the wood. The hard bast, as de Bary (4) has noticed in Cornus, is 
found only in the primary wood, except in Mastixia, where it is formed 
secondarily and stretches irregularly throughout the soft bast, with no 
definite arrangement. In Mar lea begoniaefolia, Helwingia, and Aucuba 
japonica no hard bast appears at all. Stone cells are quite common through¬ 
out the family. 
In regard to the woody axis proper, I will speak first of the vessels. 
According to Sertorius these are nearly always scalariform. In the genera 
Alangium, Torricellia , and Marie a, however, there is a single pore in 
the secondary wood.' However, the pore is often drawn out, and in the 
primary wood bars are found. In some of the above instances one or two 
bars have been found appearing even in the secondary wood. The vessels 
themselves are usually small and isolated, with a four-sided appearance 
from a transverse view. The walls, according to Sertorius, have always 
bordered pits where the vessels come in contact with the parenchyma of the 
rays, though Solereder (5) finds simple pits where Griselinia is in question. 
The wood parenchyma has only bordered pits in some cases, while in others 
it may have both kinds or only the simple pits. Septation of tracheides is 
not a common feature, though Solereder reports it for Mar lea, and I have 
also noticed it in Aucuba japonica and Griselinia lucida root. The rays are 
in general small, being from one to five cells in width. In our native species 
they are not very abundant, but in some of the exotic genera, such as 
Griselinia, Corokia , &c., the rays are very numerous, and the individual cells 
are often quite large. 
