OXALIS CORNICULATA AND ITS ALLIES 889 
both radical and conservative, were inclined to urge the importance 
of legislation on this matter during the recent Congress at Vienna, 
no action leading to a definite solution of the problem was taken. 
We are still in the dark as to which of several elements described 
as ecies shall in stints esi hi retain the old name, 
Test oee in such ¢ al ve been, and still are, widely 
divergent. The general rodles that the most characteristic part 
of a group which is to be divided shall continue to bear the ae 
name is often so vague in its application as to bring about no agree- 
ment whatever. Of this fact, the yellow-petaled species of Oxalis 
furnish a drastic example. eir distinctions have been clearly 
x 
4 most r pects ad e—may be bc dig the pro- 
cialis validity of types. 
n the meantime it is necessary to make use of such general 
a as have been locally practised by those who have given 
the subject careful thought, and have aimed to ge consistent. Of 
these principles there are at least two which possess a reasonable 
definiteness—(1) the principle of priority of position, according to 
which the first-mentioned type determines the application of “the 
name ; and (2) the principle of residues, according to which a sub- 
sequent author may remove any portio ae species or other com- 
ein, é 
two divergent principles, we may profitably apply each to the Seal 
group in hand, and see what the results will be. 
According to the rar of priority of position, O. corniculata 
L. must rest upon the plant of the Hortus een which, as 
stated above, is our Gpctics no. 2; and O. stricta L. must rest upon 
the plant of Clayton, which is our species no. 3. Finally, the 
ciel plant (no. 1), which formed the non-typical part of the 
Linnean OQ. corniculata, must receive its earliest subsequent name. 
This @ appeats to be O. repens Thunb. Oxal. 14 (1 , 
the other hand, the principle of residue is applied, the 
‘esi is as follows: (. cornic ulata o have s as & 
composite species until the publication of rienbeg, mentioned 
above. At that time (1781) re Seve, wa plant (our species no. 1) 
was taken out, ary independently described as O. repens, which 
leaves the Linnean name 0. cormentat to stand for the stoloni- 
ferous plant (our species no. 2). n t 
thus been applied to no. 2, it is svigant that the Lin 
O. stricta, originally applied to nos. 2 and 8, can #5 as cone 
no. 3. 
Thus by a curious and very happy coincidence the two different 
methods lead in this group by diverse paths to identical results. 
The three species under diseussion may therefore be collated, with 
their leading synonymy, as follows :— 
