408 W. L. Bray. 
regions could scarcely be distinguished, and although the Chilan and 
Australian regions offer essentially the same conditions for this unvarying 
growth that it finds in Capland, their corresponding species are very 
different. We must suppose therefore, that a very recent interchange 
between the regions has scarcely taken place. In what way could an 
interchange occur? The seeds of these plants show no special adaptations 
for trans-oceanie distribution. They have no means, except their minute- 
ness, by which they would adhere to the feet or legs of coast birds, nor are 
they likely to prove sources of food for such birds and thus come to a wide 
distribution as undigested particles. 
We must consider further that these species do not stand indepen- 
dently in the regions in which they oecur but must be considered a part 
of the extensive developement of the genus in those regions, so that what 
has been said concerning the close relationship between the three annual 
species applies likewise to the shrubby perennials. It has already been 
pointed out that F. laevis in the Mediterranean region becomes a transition 
form, F. Krebsii, in Capland, and that F. nothria of Capland has its nearest 
relation in F. pauciflora of Australia. Further, that these two are bound 
by very intimate characteristies to the common shrubby species of Chili. 
In faet, then, these four regions, Mediterranean, Capland, Australia, Chili, 
possess a modern developement of the genus Frankenia, each analogous, 
and nearly related to the other. This repetition of conditions in the diffe- 
rent regions points to a facility of interehange entirely inconsistent with 
the geographical isolation. As indicated, the continuous coast between the 
Mediterranean and Capland regions can account for the relation of these 
two, and the same may also be said to aecount for F. grandifolia in Cali- 
fornia and the adjacent territory. The transoceanic distribution is however 
quite a different matter, the further consideration of which may be left 
until the final discussion of the probable history of the Frankeniaceae. 
Toichogonia isolata. 
Those species which are referred to section Toichogonia isolata from 
South America are brought together upon negative characters perhaps as 
much as upon positive, they are those species which do not belong to 
Basigonia or to the Toichogonia cosmopolita groups. As previously stated, 
the one positive character lies in the floral structure, which in minuteness 
and in the method of ovule formation is like Eufrankenia. Otherwise these 
species stand sharply distinct from each other both in habit and distribution. 
They embrace provisionally, F. triandra Remy (incl. Pycnophyllum sulcatum 
Griseb. Pl. Lorenz. p. 28) from the Puna Region, F, farinosa Remy, northern 
Atacama, Cobija, placed here from the description, no specimen having 
been seen, F. Vidalii F. Phil., Islands of San Felix and San Ambrosis, F. 
