14 
Fig. 13. Tristiropsis obtusangula, impression of leaflet, 
showing texture and venation. 
Family unknown 
Unidentified impression, probably a leaf scar. 
ID2(F)-5 
This is a concave impression, more or less 
shield shaped, with 5 (or 6) bundle traces. 
Unidentified mold of a fruit or curved leaf. 
PC-86-3 
This hollow curved impression, lacking 
suggestive details, was not matched with 
anything. 
PALEOECOLOGY 
By listing the species identified from each 
of the samples it is possible to suggest some- 
thing of the probable ecological conditions 
and vegetation types at the time the various 
deposits represented were laid down. Such 
conclusions are, of course, extremely tenta- 
tive when based on so few species. It is 
greatly hoped that at some future time more 
extensive collecting may be done in these 
beds so that the picture of the past vegetation 
and conditions on the island may become 
clearer. 
PACIFIC SCIENCE, Vol. XII, January, 1958 
Fig. 14. Premna obtusifolia, an unusually small leaf 
blade, with apex missing, showing venation. 
Sample PC-43 
Pandanus tectorius , Miscanthus floridulus , Alo- 
casia macrorrhiza 
It is hard to associate Alocasia with sword- 
grass except where the Alocasia might have 
been in a somewhat shaded ravine bottom, 
perhaps with Pandanus trees, and with the 
sides of the ravine covered with swordgrass, 
possibly extending down almost to the bot- 
tom of the ravine. On more level ground, if 
it is not too dry, the Alocasia could have 
grown in a thicket of Pandanus and possibly 
other trees, closely surrounded by swordgrass. 
The Alocasia suggests the presence or former 
presence of man, as this plant was in all 
probability distributed in the Pacific islands 
entirely through human agency. 
Sample PC-48 
Miscanthus floridulus 
Probably swordgrass vegetation. 
Sample PC-58 
* T ristiropsis ohtusangula 
This is a forest tree, not now found on 
* Not now known to be living on Pagan. 
