58 
The other attempt to arrange the components of the genus was 
made by Pax and Hoffmann in 1912 (Pflanzenr. IV. 147, vi. 
pp. 18, 19). Adopting the characters relied on by Miiller, these 
authors have treated the group with 3-seriate anthers and stellate 
pubescent ovaries as a distinct section, 1. Plicatae. Those forms 
with 2-seriate anthers and lepidote capsules were subdivided into 
two more sections. Pax and Hoffmann have treated Miiller’s 
last group, in which the seeds are smooth, as a distinct section, 
2. Senegalenses; the forms with 2-seriate anthers, lepidote cap- 
sules and tuberculate seeds they have regarded as forming a third 
section, 3. Tinctoriae. As compared with the arrangement pro- 
posed by Miiller, that of Pax and Hoffmann supplies a section 
1. Plicatae, corresponding with Miiller s section 1, with two se 
tions, 2. Senegalenses and 3. Tinctoriae corresponding, in reversed 
order, with Miiller’s two groups under Section 2, sub-section 3. 
quite consistently with the group which constitutes Miiller’s 
Section 2, Subsection a. In their careful Conspectus Sectionum 
(l.c. p. 19) these authors have been at pains to exclude the group 
in question both from their sections 2. Senegalenses and 3. Tinc- 
toriae, though they have elsewhere (l.c. pp. 18 and 21) included 
it under 3. Tinctoriae, with the re a members of which the 
separates it from their section 3. Tinctoriae, when they state that 
Miiller’s second group cannot satisfactorily be included in the 
Tinctoriae, and to throw doubt on the conclusion (1l.c. p. 18) that 
this group has been derived from their section 3. Tinctoriae: 
vertently tran 
generic description (l.c. p. 17) and in the conspectus (1.c. p. 18), 
th of the sections Senegalenses and 
