Primitive A ngio sperms. 145 
this early cambium, and secondary growth is never resumed in this 
tuber. 
The bundles of the second tuber, however — that which springs from 
the tuber of the seedling — possess cambium which is particularly active 
in the second season. Professor Oueva figures the radial rows of elements 
added to phloem and xylem ( 65 , Figs. 72 and 90). He does not describe 
the formation of intrafascicular cambium at any time. 
All succeeding tubers show a similar formation of secondary tissue, 
more massive than in the second tuber (p. 56, and p. 58). 
The presence of a functional cambium in the mature bundles of 
a Monocotyledon may be advanced to show that they are derived from 
an ancestor which possessed one. Professor Queva takes this view : — 
e The persistence of a cambial zone in the bundles of certain Mono- 
cotyledons shows that we may logically consider them as derived from the 
more primitive Dicotyledons by means of the early disappearance of the 
cambium, and an increase in the number of traces from each leaf.’ (Trans- 
lated from 65 , p. 147.) 
Comparison of seedling anatomy in the two classes, then, leads to 
the conclusion that the dicotyledonous stem-structure is primitive, the 
monocotyledonous derived from it. This conclusion agrees with that 
founded on comparison of the mature stem anatomy in both classes with 
that of other groups, living and extinct. The presence of a cambium is the 
rule, its absence the exception. Except in Gloriosa , it is completely lost 
in the mature stem of Monocotyledons, but it can still be found occasionally 
in the traces of their seedlings. 
The Primitive Angiosperms, then, must be credited with a cambium, 
which has been very completely lost by one branch of their descendants. 
Number oj Cotyledons. 
Monocotyledons differ from Dicotyledons in other mature characters 
besides the structure of their stems. But these are far less constant, and 
may be considered later. An embryonic character — the difference in the 
number of cotyledons — must come first, for in systematic importance it 
ranks with the stem anatomy, or even takes place of it. 
No Monocotyledon with which I am acquainted has two cotyledons. 
Some Dicotyledons, indeed, have only one. The species with this abnormal 
feature are widely separated in the Natural System. They belong to the 
Ranunculaceae, Fumariaceae, Umbelliferae, Primulaceae, Lentibulariaceae, 
Nyctaginaceae ( 72 , p. 76). Perhaps the species of Peperomia lately 
described by Mr. A. W. Hill also belongs to this list ( 39 ). Altogether 
I cannot find quite twenty species of Dicotyledons in which this character 
has been recorded. The possession of one cotyledon then is universal among 
Monocotyledons, and the possession of two very nearly so among Dicotyledons. 
