395 
Adenoc'lirieae of South Africa. 
of Thunberg. This is quite intelligible. What, however, is not intelligible 
is that Baillon did not treat the name Adenocline Mercurialis , Turcz., as 
a synonym of A. acuta , Baill. ; he maintained the name A. Mercurialis , 
ostensibly on the authority of Turczaninow, to connote a mixture of Para- 
denocline procumbens , Mull, arg., Leidesia capensis , Mull, arg., and L. obtusa , , 
Mull. arg. The use of the name A. sessiliflora in place of A. sessilifolia is 
no doubt a mere lapsus. But it is hardly possible to explain the retention 
of A. paucijlora , Turcz., after Turczaninow had confessed that no such 
species exists. Finally A. humilis , Baill., which is said to be Mercurialis 
serrata , Meissn., and if it be so is therefore Adenocline ovalifolia , Turcz., 
cannot possibly be A. humilis , Turcz. In 1862 Baillon treated the five 
species recognized in 1858 rather differently. All of them were now looked 
upon as species of Mercurialis ; some of them only were placed in the 
section Adenocline. The A. Mercurialis of 1858 was transferred to the 
section Trismegista and subdivided into M. capensis, corresponding to the 
genus Leidesia as a whole, and M. violaefolia , corresponding to Muller’s 
proposed genus Par adenocline. The other four species were now reduced 
to three, one of which, M. caffra , is the opposite-leaved Adenocline acuta , 
Baill., the true A. Mercurialis , Turcz., as to the identity and limitation of 
which mistake is barely conceivable. The non-existent A. paucijlora was 
still kept up as M. paucijlora , while all the alternate-leaved species of 
Adenocline , Turcz., and Diplostylis , Sond., with, in addition, M. annua , 
Drege non Linn., which is Baillon’s own M. violaefolia , were lumped together 
as M. bupleuroides , Meissn., the one species described by Meissner of which 
Turczaninow never saw a specimen. 
Only one more step was required to complete the confusion. This 
MUller took when, in 1866, he united the conglomerate ‘bupleuroides’ and 
the fictitious ‘pauciflora’ of Baillon under the name Adenocline paucijlora , 
MUll. arg. (DC. Prodr., xv. 2, 1139), thus using for this widened species 
the one name employed by Turczaninow in 1843 which that author in 
1852 found it necessary to cancel. That the species thus proposed is 
a composite one needs no demonstration ; MUller has tacitly admitted the 
fact by making a careful attempt to separate as varieties the species of 
Turczaninow, Sonder, and Meissner. In so doing, MUller has not always 
succeeded in keeping clear of the pitfalls dug by his predecessors ; he has, 
besides, initiated difficulties. Among these may be noted the statement 
under y humilis that it is the female portion of Drege 3441 which is the 
same as A. humilis , Turcz., whereas we know from Turczaninow’s statement 
and from the actual specimens that the female part of Drege 3441 has 
opposite leaves. MUller has referred Drege 1867 both to d bupleuroides and 
to C transients', the statement that Hout Bay, mentioned under e serrata in 
connexion with Krauss 1190, is in Natal is erroneous — the locality is in the 
Cape Peninsula ; the statement under 6 tenella that Krauss 1191 is a Natal 
