1234 LEAFLETS OF PHILIPPINE BOTANY [Vor. IV, Arr 67 
and gall flowers in one set of receptacles; fertile female 
flowers in another set. .. I. Palaeomorphe 
Group IHi.—Unisexual or asexual; male flowers without 
rudimentary pistils. 
Section I.—Male, gall and fertile female flowers on the 
same. receptacles. 5 TE. Uxostigma. 
Section ll.—Flowers unisexual or neuter; male and 
gall flowers in one set of receptacles, fertile female 
and neuter flowers in another set... III. Synoecia. 
Section III.—Flowers unisexual; male and gall flowers 
in one set of receptacles, fertile female flowers only 
in another set. 
A.—Flowers monandrous. 
a. Receptacles chiefly axillary....IV. Sycidium. 
b. Receptacles mostly in‘fascicles from stem 
and: branehes.. |... V. Covellia. 
B.—Flowers di rarely triandrous: 
a. Receptacles mostly axillary... VI. Eusyce. 
b. Receptacles mostly in fascicles from stem 
and branches... VII. Neomorphe. 
By this key it is exceedingly difficult to bring a species into 
its rightful section. Our specimen may or may not have stam- 
inate flowers, or vice versa. Dr. Warburg has already complain- 
ed of the sectional characters which no doubt led him into a few 
errors. The writer's own limited study of Philippine Ficus has 
convinced him that there are no divisional characters between the 
monandrous species with receptacles mostly in fascicles from 
stem and branches and the di or triandrous species with recep- 
tacles mostly in faseicles from stem and branches. Again, the 
fig flowers are not easily classified by the amateurs, the stamens 
of certain species are hard to find or even to recognize when 
they are present. It is no easy matter to detect the young fertile 
female flowers from the neuter or from the pseudo-hermaphro- 
dite flowers. So I have come to feel that if even a less scientific 
or natural classification could be evolved from the external ap- 
pearances it would prove no less serviceable. In the following 
rough sketch I have outlined such a course for most of our Phil- 
» ppine species, and with further construction it may become 
at least artifieially useful to systematists. In glaneing over it 
Local 
