FLORAL ANATOMY OF SOME COMPOSITAE. 517 
On the Floral Anatomy of some Composite. By James SMALL, M.Sc.(Lond.), 
Lecturer in Botany, Durham University. (Communicated by Prof, M. 
С. Porter, M.A., Sc.D., F.L.S.) 
(4 Text-figures.) 
[Read 80th November, 1916.] 
THERE have been numerous observations on the venation of ihe corolla in the 
Composite (2, 4, &c.) ; the vascular supply of the ovule (6, 8, 14, 18, 19) 
and the style (5) has also been the subject of enquiries, and there are also 
records of the structure of the pericarp in many species (3, 6, 7, 9); but, 
except in the Cichoriew (15, 16, 17), no complete study of the vascular 
system of the flower appears to have been made. The observations of Trécul 
confirm the striking uniformity of the vaseular supply of the corolla in the 
Cichorieæ, which seems obvious from the external examination of the 
venation of a large number of species and genera belonging to this group. 
The present investigation deals with the floral anatomy of three typical 
Forms of florets—tubular, ligulate, and bilabiate or pseudo-ligulate. As it 
has been suggested that the Cichorieæ were derived from the Senecioneæ 
(10, 12), it was thought that the examination of the linear ray florets in Tus- 
silago Farjara might show an intermediate stage between the two groups. 
The anterior lip of the labiate or ray floret in the Composite is frequently 
four-lobed, especially where the corolla is broad, as in some species of Creman- 
thodium (С. rhodocephalum, Diels, C. Decaisnei, C. B. Clarke, С. reniforme, 
Benth., C. Thomsoni, С. B. Clarke). The vascular supply is modified 
accordingly. The corolla may be broad and still only three-lobed, as in 
Layia elegans, Torr. & Gray, or it may be three-lobed with a vascular 
supply for four lobes, or the lobes may be fused so that the number varies 
from one to three, while the number of conducting strands remains the same. 
When the anterior lip is broad the conducting strands tend to increase in 
number, the simplest modification being the development of three auxiliary 
strands in the position of the midribs of the three petals. Branching of the 
strands may take place to a varying degree, and usually occurs in the above- 
mentioned species of Cremanthodium and in many Senecio species, as well 
as in many other cases where the number of main strands remains three or 
four or by fusion becomes less than three. The number of strands in 
S. Doronicum, Linn., varies from four to eight, and in S. eubœus, Boiss. & 
Heldr., may be as many as eleven, but it may be larger in other species 
(e. g., Helianthus spp.). 
The vascular system of the ovule has been the subject of many obser- 
vations and some controversy, but it does not seem to have any bearing on 
the problems of phylogeny. The occurrence of anomalous biovulate and 
LINN. JOURN.—BOTANY, VOL. XLIII. 2.0 
