210 
catalogue of his herbarium, which appeared in 61 different parts 
between 1791 and 1827. The main list of the herbarium was 
published in 1791-1797, and the part treating with the class 
Decandria is of the year 1793. This part enumerates only three 
of his Dianthi from the Cape (crenatus, cespitosus, scaber). It is 
only in an Appendix of the year 1806 that we find Dianthus 
albens. The name D. incurvus is to be found nowhere in his 
catalogues. : 
“These facts are corroborated by consulting his manuscript 
catalogue, which essentially corresponds to his printed catalogue 
of 1791-97, and in which Thunberg has added in the margin the 
species enumerated in his Appendix of 1806 mentioned above. 
Here the name “ albens”’ 
the margin. 
‘From these facts it seems probable that Thunberg at the 
time of his Prodromus disposed of no specimen identified by him 
with D. incurvus. The description in the Prodromus might have 
been made in S. Africa. But later he seems to have found 
is to be found among the species in 
fore wrote ‘“‘ Dianthus albens, Wild.” and added as a synonym 
* D. incurvus, Prod. cap.” 
In the Flora Capensis, Sonder recognised nine species of 
Dianthus. Of these D. incurvus, Thunb., and D. holopetalus, | 
Turez., prove inseparable, as also do D. prostratus, Jacq., and 
D. pectinatus, E. Mey., thus leaving seven valid species in the 
Flora Capensis. To these must be added :-— 
D. micropetalus, Ser. (1824), placed by Sonder under D. scaber, 
hunb 
unb. 
D. Burchellu, Ser. (1824), placed by Sonder under D. incurvus, 
Thunb. 
~  D. namaensis, Schinz (1897). 
D. mootensis, Williams (1889). 
In the present paper six additional species and three varieties 
are described for the first time, bringing the total number of 
South African Dianthi up to seventeen species and three varieties. 
The rich material now available at Kew, shows that the 
simple or branched habit of the flowering-stem, used by Sonder 
to group the South African forms, cannot be relied upon, even 
as a specific character, many individuals bearing both simple 
and branched fiowering stems. 
Much of the difficulty experienced by authors in placing some 
of the material with certainty, has been due to poor preparation 
by collectors. In this genus the cutting of the petal margins is 
of some diagnostic value, yet many specimens have been dried in 
such a way that this character cannot be distinguished. The relat- 
ive size of the basal and intermediate cauline leaves, the relative 
