136 BOTANICAL GAZETTE [FEBRUARY 
theory is based on the study of the seedlings of a large number 
of plants, many of which show the cotyledons grown together 
at one or both sides. 
To determine which of these theories is the more probable 
will require the study of the embryology and development of the 
seedlings of many plants in the families which show these pecul- 
iarities. Something of value may be learned by growing the 
seedlings under different conditions and observing results. 
The interpretation of such a structure as that found in Podo- 
phyllum, or that of Nelumbo as one cotyledon or as two, must 
be unsatisfactory until a greater number of anomalous forms has 
been investigated. If the cotyledons of some of the plants 
which have the other characters of the dicotyledons are found 
to have their origin as a single ridge, it will support the view that 
here also we have two cotyledons and not lobes of one. The 
early division of the ridge in Podophyllum, as well as in Jeffer- 
sonia and Caulophyllum, leads to the conclusion that here we | 
have two cotyledons. The fact that a cotyledonar tube is formed 
in Podophyllum is probably best explained as a derived con- 
dition brought about by the geophilous habit of the plant. The 
hypocotyl is short in comparison with the length of the entire 
embryo. The plumule is small and develops but little during 
the first year. The cotyledonar structure is long, so that, while 
the plumule remains underground in a protected position, the 
cotyledons are pushed up into the air and 
sunlight and carry 0m 
the work of assimilation. 
The great length of the cotyledons 
the tube. The fact that cotyledonar t 
families far removed from each other, and usually in embryos 
having a short hypocotyl, Supports this view. 
In Caulophyllum the ridge does not attain much height, and 
the later stages show that the notches 
ubes have been found in 
lum. Possibly this explains why no tube j 
It is highly desirable to trace the development of the seed- 
