A denoclineae of South Africa. 385 
other — these two plants so closely resemble each other that even in the 
field the most careful of collectors have at times been led astray. 
To Muller our thanks are due for having, in 1866, determined this 
double confusion. He rectified matters by establishing the genus Leidesia 
for the reception of M. capensis, Sond., and at the same time by distin- 
guishing as L. obtusa the somewhat elusive species which, as to bibliography, 
had hitherto been merged in its 2-coccous congener, but, as to practice, 
was in herbaria mainly to be met with in the covers containing the 3-coccous 
South African species with flaccid, pellucid leaves. This last species, with 
androgynous inflorescences, 3-coccous capsules, 2 -partite styles, and imbri- 
cate calyx-lobes, Muller treated as a distinct genus, Paradenocline , Mull, 
arg., most nearly allied to Adenocline , Turcz. When doing so, Muller 
pointed out that these two genera, Adenocline and Par adenocline , conjointly 
form a distinct and natural group, which he named Adenoclineae and 
referred to the sub-tribe Hippomaneae (DC. Prodr., xv. 2, 1139). 
The recent history of these South African groups has not been eventful. 
It has, however, been found necessary to add to Leidesia a new species, 
L.firmula , Prain (Kew Bulletin, 1912, 337), and to Adenocline another new 
species, A. stricta , Prain ( 1 . c., 338). 
Review of Seidelia. 
The genus Seidelia was established by Baillon in 1858 (£tud. gen. 
Euphorb., 465, t. 9, fig. 7) for two South African species till then included 
in Mercurialis. One of these, first collected by Drege in the Orange-Vaal 
basin, is M. triandra , E. Mey. (Linnaea, iv. 237), published in 1829; the 
other, found by Zeyher in the Coast Region of Cape Colony, is M. pumila, 
Sond. (Linnaea, xxiii. 112), published in 1850. In 1862 Baillon replaced 
Seidelia , as a distinct section, in Mercurialis (Adansonia, iii. 160, 175). 
When dealing with Seidelia in 1866 (DC. Prodr., xv. 2, 947) Muller 
differed from Baillon in two respects. Muller treated the two species 
accepted by Baillon as varieties of one ; at the same time he transferred the 
whole section ‘ Seidelia ’ from Mercurialis to Tragia . The only apparent 
reason for this transfer is that Seidelia and Tragia normally have each 
three stamens. The transfer was not accepted by Hooker in 1868 (Harv. 
Gen. S. Afr. PL, ed. 1, 338) ; it was repudiated by Bentham in 1880 (Gen. 
PL, iii. 311). Hooker followed Baillon in recognizing two species ; Bentham 
accepted MiillePs view that there is but one variable Seidelia. For this 
view there was as little ground as there was for the transfer of Seidelia to 
Tragia . Bentham has stated that Seidelia is closely allied to Adenocline ; 
the differences between the two as regards flowers, to which Bentham was 
fully alive, make this statement as hard to accept as Muller’s view that 
Seidelia is closely allied to Tragia. There is, no doubt, a general similarity 
between some species of Adenocline and certain species of Mercurialis \ 
