A Comparison of the Stem Anatomy of the 
Cohort Umbelliflorae. 
BY 
CARL S. HOAR. 
With Plates IV and V. 
A S classified at present, the families Cornaceae, Araliaceae, and Umbelli- 
ferae are placed together under the cohort Umbelliflorae. This cohort 
is in turn placed at the top of the sub-class Archichlamydeae, next to 
the sub-class Metachlamydeae. Such is the classification given by Engler 
and Prantl and followed by Engler and Gilg (1). 
The closeness of relationship between the families of this cohort has 
always been more or less in doubt. Because of this doubt I have made, 
during the last year, a special study of the stem and root anatomy of the 
three families. My purpose in so doing has been to learn what additional 
evidence I might obtain as to their true relationships. In pursuing this 
study I have concerned myself especially with the Cornaceae, having been 
able to obtain all of the genera and many of the species native to this 
region, together with several which are of exotic origin. With the other 
two families I have had a smaller number of examples, but their constancy 
of common characteristics seems to confirm my conclusions. 
Of the three families, the Cornaceae are generally considered to be the 
lowest, while the Araliaceae are placed between the former and the Umbelli- 
ferae. Such a view might easily be drawn from external characters alone. 
The Cornaceae are, with one or two exceptions (Cormis canadensis ), shrubs, 
which sometimes, as in the case of Nyssa sylvatica , become tree-like in 
appearance. The Araliaceae would seem to stand in an intermediate 
position, since, although they are often shrubby and even tree-like, yet the 
wood is seldom as well developed, and many of its members are herbaceous. 
Finally, the Umbelliferae would seem to be the highest since its members 
are always herbaceous, with very little wood. It has been conclusively 
shown that herbaceous plants are derived from the woody plants, frequently 
by the formation of large rays and the decrease of the amount of woody 
elements (2, 2 a ). For this reason the external appearance of the three 
families gives the basis for the conclusions which are shown to be true 
when the flower and stem anatomy are taken into account. 
[Annals of Botany, Vol. XXIX. No. CXIII. January, 1915.] 
