Campbell—Studies on the Araceae. 21 
less conspicuous ; the root in Lysichiton seems to be of lateral 
origin, as it is in Pis da. 
8 . A suspensor is never developed, this being associated 
with the complete investment of the embryo by the endosperm. 
Owing to the small number of types of Araceae which 
have been investigated, and the incomplete knowledge of 
most of those which have been studied, it is not yet possible 
to be very certain as to the conclusions to be drawn from 
these studies. 
Engler 1 considers the forms like Pothos and Anthurium , 
with hermaphrodite flowers, to be the primitive forms of which 
those with unisexual flowers are reduced forms. It seems 
quite as likely that the reverse is the case, especially as 
Anthurium and its allies have more complex pistils than 
most of the unisexual forms. Moreover, from comparison 
with other low Monocotyledons, it seems probable that the 
forms with a single carpel and single basal ovule are the most 
primitive. However, we are not in a position yet to decide 
this important question. 
The development of the embryo-sac requires much more 
attention. In Arisaema triphyllum there are regularly four 
archesporial cells resulting from a longitudinal division of 
the primary archesporial cell 2 , and it remains to be seen 
whether this may not be the case in other forms, instead 
of the simpler type found in Dieffenbachia. 
The early development of a solid endosperm seems to be 
a pretty constant character in all the forms yet examined, 
and is an important one. A comparison with the prothallial 
tissue of Isoetes or Selaginella is inevitable, and it probably 
represents a primitive condition as compared with most 
Angiosperms. 
The great development of the antipodal cells is also to 
be considered as an evidence of the primitive nature of the 
1 Engler and Prantl, Natiirliclie Pflanzenfamilien. 
2 Mottier, 1. c. 
