The Geographical Distriaution of the Frankeniaceae ete, 415 
It was mentioned earlier as a notable circumstance that no species of 
Frankeniaceae occurs on the Atlantic coast of North or South America 
(except the introduced F. pulverulenta in New York Harbour). Likewise 
noteworthy is the lack of any representative of the family in the Eastern 
Asiatic salt deserts, or on the east Asiatic coast, so that a distribution from 
the Mediterranean region to the Western Hemisphere has probably not 
occurred. Much more is it likely that the distribution has proceeded from 
Australia to the American continent reaching naturally, the nearest region, 
the Chilan coast, first, which is born out by the strikingly close relationship 
of F. serpyllifolia Australia to F. grandifolia Chili and of F. pauciflora 
Australia to F. chilensis, erecta, glabrata etc. of Chili. 
It seems also probable, from the very extensive developement in 
Australia that F. nothria of Capland is of Australian origin, and that in the 
Capland region we have the meeting of Australian and Mediterranean 
elements. 
Whether now the original center from which all of this modern devel- 
opement arose was the Mediterranean region or the Australian, is not 
easily determined, if at all possible. 
Three facts speak in favor of a southern extratropical origin. 
1. The lack of Frankeniaceae on the East American coasts or in the 
East Asiatic salt-steppes. 
2. The predominance of only the closely related species of the modern 
developement in the Mediterranean region, and conversely. 
3. The abundance of monotypic isolated forms in the southern extra- 
tropical regions. 
Whatever may have been the direction of distribution, we must 
conclude that by some means, seeds of Frankenia have been carried over 
the regions between Australia and the Capland, and between Australia and 
the western South American coast. 
If the previous interpretation of the relation of species be correct, a 
distribution by means of shipping would not allow sufficient time for the 
varied developement within the regions. 
It is impossible that such a distribution could be made by means of 
ocean currents, if for no other reason than that the seeds themselves are 
tiny grains which quickly sink in water. 
Further, the seeds are in no way adapted for clinging to objects, nor 
are they apparently food for any birds which are coast frequenters and at 
the same time birds of long flight, and yet there is this small chance, that 
in some accidental way they might remain attached to the body or in the 
alimentary canal of one of those birds which are able to remain on the 
wing for days at a stretch, and could, so far as the mere ability is concerned, 
make the flight between Australia and the Capland or Chili. 
That other plants have attained a distribution over those same stretches 
