744 
Lee. — Observations on the 
Cactaceae, the writer inclining to the view that the two groups ‘ are derived 
from the same ancestral stock ’, and that ‘ it is no longer necessary to look 
upon [the occurrence of the Anemarrhena type in the Cactaceae] as acci- 
dental. Among the Cactaceae this type of transition occurs in those genera 
whose seedlings are least modified, and it is in such cases that we should 
naturally expect to find that ancestral characters had been retained.’ The 
conclusion of A. Arber is ‘ that Miss Sargant’s view that the various mono- 
cotyledonous and dicotyledonous types of seedling anatomy were all 
originally derived from the Anemarrhena type is strengthened, rather than 
weakened, by the discovery of the ancestral type among the Cactaceae.’ 
In making this criticism the writer admits that the question depends 
very largely upon the systematic position of the group in question. The 
force of the criticism is, in fact, proportionate to the real distance between 
the Cactaceae and the group from which it was derived, which is assumed — 
an assumption that is itself not unquestioned — to have given origin also to 
the Nymphaeaceae and the other members of the Ranalian plexus. If, 
therefore, there exists a group which shows specialization in the ‘ parts 
of prime importance as regards affinities ’, i. e. ‘ in the reproductive organs, 
and not the vegetative features of the mature plant ’, and which also possesses 
the Anemarrhena type of seedling structure, it will perhaps be conceded 
that some ground exists for the view that the resemblance between the 
seedlings of this group and those of E rant his , &c., is accidental. Such 
a state of affairs is found in the Bignoniaceae. Even assuming that this 
order was originally derived from the same group as the Ranales, there can 
be no doubt as to the want of affinity between the two, and that any 
character which is common to Incarvillea Delayvei (Bignoniaceae) and the 
Ranales serves to indicate nothing more than an analogy, and is, to use 
de Fraine’s term, quite ‘accidental’. Neither can there be any reasonable 
doubt that the discovery of the Anemarrhena type in this highly evolved 
group has not only ‘ considerably weakened the evidence on which the 
“ fusion ” hypothesis was based ’, but also has placed ‘ Sargant’s view that 
the various monocotyledonous and dicotyledonous types of seedling 
anatomy were all originally derived from the Anemarrhena type ’ beyond 
the region of probability. 
It is highly probable that many other species, more or less remote 
from the Ranunculaceae or Liliaceae, will be found to possess the Anemar- 
rhena type of seedling anatomy. The observations on Convolvulus 
recorded above indicate the absence of any sharp line of demarcation 
between the Anemarrhena method and Van Tieghem’s Type 2. What is 
practically the latter is seen in Convolvulus tricolor , var. major . If, in this 
example, the lateral cotyledonary strands fused with the median one, and, 
at a lower level, separated and became orientated to form the tetrarch root, 
the Anemarrhena type, with, however, two cotyledons instead of one, would 
