69 
C. Robertson* has ably discussed the question of the primitive 
or non-primitive character of the ‘‘ Amentiferae”’ group o 
Apetalae, on which the acceptance of Engler’s views so much 
depends. He offers the theory that primitive Angiosperms were 
already entomophilous and that anemophilous types are meta- 
morphosed entomophilous flowers resulting from reduction and 
amas and are not at all primitive. Probably too much 
importance has been attributed to the anemophily of the 
Ainenssfori, on the false | analogy of the Gymnosperms. 
Regarding the origin of Ranunculaceae and allied families, 
after a careful examination of these groups, I find myself unable 
to accept Hallier’s view} that the Ranunculaceae and Nymphae- 
aceae have descended from the Magnoliaceae through the 
Schizandraceae, Lardizabalaceae and Berberidaceae. It is a far 
cry indeed from a ueTOND to a Magnolia. To my mind the 
floral structure of Berberidaceae and Lardizabalaceae suggests 
evolution in caiely the opposite direction, i.e., from the 
Ranunculaceae. There are many connecting links between the 
Ranunculaceae and the Berberidaceae, and the latter appear to 
be the reduced and more recent forms. The relative antiquity of 
woody and herbaceous plants has been the subject of a most 
interesting and instructive paper by Sinnott and Bailey, whose 
. researches go to show that in the flowering plants herbs, as a rule, 
have been derived from woody plants. However probable this 
may be, I do not think it can account for the origin of Ranuncu- 
laceae from any existing woody group. That herbs have been 
derived from ligneous types in certain families seems clear from 
some of the examples they quote, but it seems to me that the 
theory should be applied with considerable caution, for it is 
reasonable to suppose that the reverse may have ‘frequently 
occurred, and that an herbaceous group may have evolved woody 
or semiwoody types. The study of this subject from both 
standpoints would no doubt be productive of interesting results. 
The type of wood peculiar to Clematis, Berberis and the Meni- 
spermaceae generally, may owe its peculiarity to derivation from 
herbaceous ancestors. From a consideration of floral structure 
and wide distribution the genus Clematis with its woody habit, 
opposite leaves, its induplicate-valvate sepals and apetalous 
condition, in conjunction with this view regarding the origin of 
woody structure, may be traced back through the intermediate 
genus Clematopsis (Kew Bull. 1920, a to the ama 
and more primitive genus Anem by 
§Pulsatilla. There are wide gaps in ti the relationships of the 
Robertson, The Structure of the Flowers and the Mode o 
Pollination of the Primitive Angiosperms : Bot. Gaz. 37: ooecton (1604). 
{ H. Hallier, “ Provisional Scheme ‘of the Natural (Phylogenetic) 
System of Flowering Plants,” in New Phytologist, 4: 151-162 (1905); 
see especially p. 157. 
E. W. preted & I. W. Bailey, “The Origin and Dispersal of Herb- 
aceous Angios: ”? in Ann. Bot. 28: 547-600 (1914). 
