64 
SOUTH AFRICAN SPECIES OF 
though in dried specimens it is sometimes almost impossible to separate 
them. It is closely allied to Gr. pharnaceoides, C. A. JVIey., the type 
of which occurs in Abyssinia. 
20. Grassula campestris (E. et Z.), Harv. in Harvey and Sonder, FI. Cap. 
ii, 351. (Plate v, fig. 17.) 
? Gr. campestris, Endl. in Walp. Rep. II, 253. 
Tetraphyle campestris, E. et Z. Enum. no. 1873, 294. 
T. lanceolata , E. et Z. Enum. no. 1874, 294. 
T. muscosa, E. et Z. Enum. no. 1872, 294 1 . 
Annua herbacea fiaccida. Caulis filiformis e basi ramosus ramis 
totum foliosis decumbentibus vel adscendentibus inferne saepius 
cum ramulis abbreviatis. Folia connata imbricata vel remota ovato- 
lanceolata vel lanceolata acuminata vel capillari-cuspidata. Flores 
pentameri in fasciculis densis axillaribus subsessiles. Calyx quam 
corolla longior lobis subulato-lanceolatis acutissimis. Petala basi 
breviter connata alba lanceolata acuminata vel capillari-cuspidata. 
Stamina quam petala breviora filamentis subulatis antheris subro- 
tundis. Carpidia staminibus subaequilonga ovariis oblique subovatis 
2-ovulatis stilis brevibus subulatis. Squamae spathulatae. 
The plant is frequently 2 — 4 cm. long, rarely more than 7 '5 cm. 
The following measurements of the floral parts were taken from 
a specimen collected at Port Alfred : Sepals about 2 mm. long. 
Petals about 1 mm. long. Stamens '75 mm. long. 
This is a somewhat variable plant which at first sight very 
frequently looks exactly like Tillaea muscosa, Linn. In fact, I have 
had specimens lying side by side from France, from near sea-level in 
South Africa, the Andes of Peru (alt. 12,000 — 13,000 ft) and numerous 
other places, and these plants could not possibly be distinguished 
without dissection of the flowers, and as these are extremely delicate 
and Herbarium material not always quite suitable or scanty, we are 
faced with a great difficulty in deciding whether Tillaea muscosa, L. 
is so widely spread as is often assumed. This species has been called 
Crassida muscosa by Roth in Enum. PI. Phan. Germ. I (1827), 994. 
I hope shortly to deal with all the species of Tillaeoideae not found in 
South Africa and further light may then be thrown on the question, 
but here I would only refer to a couple of points which make it 
desirable to suspend judgment and influence us, at all events for the 
present, in retaining the name which Harvey has adopted. 
Tillaea muscosa, L. in Sp. PI. (1753), 129, has been frequently 
described and figured, e.g. by De Candolle in PI. Grasses, t. 73. In 
1 A critical discussion of the possible further synonymy is being prepared, and will be published 
separately. 
