98 KEVISION OF INDIAN SCREWPINES AND THEIR ALLIES. 
hemisphere. Their centres are in the Malayan archipelago and in Mada- 
gascar. Their northern limit is in China and in the Himalaya hills, 
where the Korr (P '. farcatus) occurs in abundance. This extension, so 
far to the north through the hilly parts of Assam, is owing, without 
doubt, to the moist condition of those countries. From Japan, too, Zuc- 
carini notes species of Pandanea (Miinchner Gelehrte Anzeigen, 1341 
and 1844, "Notizen liber die Flora von Japan u. die bisher hieriiber 
vorliegenden Leistungen"). (This work is not accessible to me.) 
In N.E. Australia and the Pacific islets we have perhaps their S.E. 
and E., in Western Africa, surely their western limits. Pandemia 
verus, Eumph., reaches the north-western Himalaya, but surely only 
cultivated, as is the case in Arabia. 
Overlooking the whole area, where Pandani grow wild, we become 
soon aware that all kinds are confined to moist regions of the tropics 
and subtropics. They prefer principally marshy forests, like those 
which are so characteristic of Sumatra and other Malayan islands. 
Pandanus verus, however, prefers more the sandy beaches along the sea. 
No species however is ever observed in Mangrove swamps. 
It is remarkable that of some species, though abundantly occurring, 
only the males, of others, only the female plants have become known. 
We see often whole tracts of the sandy seashores covered by female 
plants of P. verus, without a single male amongst them ; so, on the 
contrary, all known plants of P. Icevis are males. Pandanm dubius 
produces fruits, which germinate quite freely, though without being 
fecundated by males. 
It might be objected that the fecundation has been effected by some 
other allied species, but the young plants do not differ at all from their 
parents. 
Hybrids are, so far as I am aware, not yet observed anywhere be- 
tween the tropics, except in gardens. Dr. Wallich, treating on the 
genus Hedychium in Hook. Journ. of Botany, v. (1853), has already 
made the same observation, and if we compare the distribution of 
hybrids in their wild state, then it is indeed striking to see the same 
restricted to the most cultivated parts of the globe, to Europe only. 
The number of hybrids there increases yearly. 
Only between different species of Blumea (B. lacera, oxyodonta, bi- 
folia) I suspect hybrids in cultivated grounds, but as yet 1 cannot 
confirm this by direct observation. - 
