52 
been described in such a way as to suggest that 
development of papillae on epidermal cells re- 
mote from stomata closely parallels that on cells 
immediately associated with stomata. This is un- 
doubtedly true for most Pandanus species, but 
there is evidence in Freycinetia excelsa and in 
an unnamed species of Pandanus cultivated in 
Florida that epidermal papillae may proliferate 
quite independently of stomata. In the former 
example some shoots had leaves with a distinctly 
papillose abaxial epidermis, although the stom- 
ata remained quite unspecialized. Clearly, there 
may be independence between epidermal and 
stomatal papillae in some Pandanaceae, and my 
account with the emphasis on a linear correla- 
tion between the two does not tell the whole 
story. 
DISCUSSION 
Patterns of anatomical features in the leaf 
which are of diagnostic value have not yet 
PACIFIC SCIENCE, Vol. XIX, January 1965 
been demonstrated in the Pandanaceae. Stomatal 
structure, which shows most variation, must be 
used cautiously for systematic purposes in view 
of its variation in a single individual. So far 
there is little indication that stomatal type is 
taxonomically significant. 
Stomata in Pandanaceae vary in a way which 
might suggest a relation with environmental 
aridity. It is clear, however, that the degree of 
stomatal elaboration is under genetic control and 
is not wholly a consequence of external environ- 
ment. Evidence is provided by a number of spe- 
cies of Pandanus cultivated side by side under 
identical circumstances in Florida. In this col- 
lection the whole range of stomatal types in 
Classes 1-5 has been seen, so that in this col- 
lection the release of stomatal potential is not 
affected by environment. Conversely, material of 
P. dubius from contrasted environments, sup- 
plied by Dr. B. C. Stone, showed no variation in 
stomatal elaboration. 
If it is probable that stomatal elaboration is 
LEAF NUMBER 
Fig. 52. Diagram representing increasing elaboration of stomata on a single shoot, the ultimate potential 
differing in different species (cf Fig. 1). 
