98 
Journal of Agricultural Research 
Vol. II, No. 3 
the commonly used citrous stocks are not well adapted. It is not impos¬ 
sible that the desert kumquat, being adapted to grow in desert soils, 
which are usually more or less saline, will prove able to withstand more 
“alkali” in the soil than the Asiatic species of Citrus, which are indeed 
very sensitive to salty soils or water. 
Being different from Citrus in so many visible characters, it is possible 
that Eremocitrus will also differ physiologically and prove resistant to 
some of the many fungous diseases that attack citrous stocks. 
NEED FOR TAXONOMIC STUDY OF THE WILD RELATIVES OF 
CULTIVATED PLANTS 
The bringing to light of the true relationships and possible uses of 
Eremocitrus, which, although described 66 years ago, has remained 
to this day practically unknown to botanists and horticulturists, is 
another link in the chain of arguments going to prove that a better 
knowledge of the wild relatives of our crop plants is indispensable as a 
preparation for their improvement by breeding. 
It is certainly surprising that a plant so remarkable as Eremocitrus 
glauca, closely related to our cultivated citrous fruits and bearing edible 
fruits in a wild state, the only desert plant known in the whole orange 
subfamily, and the hardiest of all the evergreen species, has never before 
been introduced into culture or utilized in breeding experiments. 
Probably the neglect of this remarkably interesting plant in the past 
has been due largely to the unfortunate nomenclatorial history of the 
species. Originally placed in Triphasia, it would naturally be supposed 
to be similar to Triphasia trifolia (Burm.) Wilson, common in gardens 
in tropical and subtropical countries. This plant is a small shrub with 
trifoliate, almost stalkless, leaves, subtended by paired spines. The 
fruit is a little berry filled with a sweetish and aromatic mucilaginous 
pulp, very unlike an orange and not at all closely related to the genus 
Citrus. 
After being classed as a species of Triphasia for 15 years, the Austra¬ 
lian desert kumquat was removed to the genus Atalantia by Bentham in 
1863. It is true that this placed it in a genus more closely allied to 
Citrus than is Triphasia, but about this time Baillon’s view that Citrus 
is closely related to the Bael fruit (Aegle marmelos (E.) Correa) and 
the wood-apple of India began to be accepted generally by botanists, 
and as a result the possibility of a species of Atalantia being closely 
allied to Citrus seems not to have occurred to any of the botanists who 
have published concerning this group of plants during the past third of 
a century. 
That the discovery of Eremocitrus is not a unique result of some 
extraordinary good fortune is shown by the fact that an equally striking 
and equally misunderstood new genus, Citropsis (Swingle and Keller- 
