eighteen inches above the soil-level and attains a height, 
with the branches, of some three feet. Burchell, who 
collected it nearly a century later in the Donderhoek 
Mountains near Villersdorp in northern Caledon, found it 
to be there very common, reaching a height of five to 
seven feet and resembling closely P. mellifera, Thunb., 
a native of the Coast Region of South Africa. The area 
occupied by P. longifolia covers practically the whole of 
the Caledon division and, if we include the form with 
smaller leaves and heads, which Mr. Phillips separates as 
var. minor, the species extends into the adjoining portion 
of the Bredasdorp division. Whether it were introduced 
in the time of Boerhaave or not, we know that towards 
the close of the XVIII. Century it was raised at Schén- 
brunn, from seeds sent or brought there probably by the 
Austrian collector Scholl. Messrs. Lee and Kennedy, of 
Hammersmith, received from Schénbrunn a plant or 
plants which flowered in their nursery early in 1801, 
and were figured by Andrews as P. longifolia nigra in 
the “Repository.” But the same firm seem to have 
obtained, also from Schénbrunn, other specimens which 
likewise flowered in 1801 and were figured by Andrews 
as P. longifolia ferruginoso-purpurea. About the same 
time Hibbert had yet another specimen, likewise sent 
from Schénbrunn, which Andrews described and figured 
as P. longifolia, var. cono turbinato. But while Andrews 
was satisfied that all three were forms or varieties of 
P. longifolia, Sweet regarded them as distinct species 
which he named P. longifolia, P. ligulaefolia and P, um- 
bonalis respectively. A fourth figure purporting to 
represent P. longifolia was published in the ‘* Botanical 
Register” in 1815. The plant in this case was again 
one belonging to Messrs. Lee and Kennedy ; this time, 
however, it was one introduced by Masson in 1790. 
Sweet accepted Masson’s plant as true P. longifolia, but 
Phillips, relying on certain differences exhibited in t. 47 
of the “ Register,” has treated it as the basis cf a dis- 
tinct species, P. ignota. In the ‘Flora Capensis”’ 
P. ligulaefolia and P. umbonalis are similarly recognised. 
As no specimens of P. ignota, ligulaefolia or umbonalis 
appear to have been preserved when their respective 
plates were published, this course is possibly safer than 
