PIIYLOGENY OF THE ARAOEAE. 
By James Ellts Gow. 
The study of the Morphology of the Aroids has been almost completely 
neglected by botanists. This is unfortunate, since the group is, in a 
sense, pivotal. Its anatomical characteristics indicate a relationship 
with the Naiadaceae and their allies; together with, a more distant con- 
nection, on the one hand with the Screw-pines and Palms, and on the 
other with the grasses. Points of relationship to the Liliaceae are not 
lacking, and there are some indications of affinity with the more primitive 
members of the Dicotyledons. At the same time, the group is by no 
means a doubtful assemblage, but has a distinctive character of its own. 
In view of these facts, its Morphology and Phylogeny should prove 
most interesting, and. should throw some light on the more general prob- 
lems of descent and relationship. 
In the past but little work has been done along this line. In 1892 
Mottier published an account of the embryo-sac in Arisaema triphyllum. 
Duggar, in 1900, described the development of the pollen grain in 
Symplocarpus and Peltandra. Between 1900 and 1905 Campbell pub- 
lished a series of articles on the Morphology of Lysichiton kamschat cense, 
An thurium violaceum , Nephihytis lib erica, Dwffenbachia seguine, and 
Aglaonema commutatum. Practically no other work has been done 
along this line up to the time when the present investigation was under- 
taken. The few scattering references to Aroids contained in occasional 
papers on closely related subjects embrace nothing of value in the pres- 
ent discussion. 
In 1906 the writer undertook the investigation of the embryo-sac of 
*a number of species of Aroids, with the intent of ascertaining whether 
s the members of this family follow the typical eight-celled plan of embryo- 
sac formation, or vary from it. As the investigation proceeded its 
scope broadened. It was found possible, in the case of many species, to 
secure good sections showing phases of megaspore and microspore de- 
velopment, and in all such cases, this was included in the scope of the 
investigation. The development of the endosperm and embryo in a 
number of species was also studied somewhat in detail, and some definite 
11 
