Holm. — On Obolaria virginica. 373 
In regard to the corolla, it may be said that this is regular, 
campanulate, and deeply four-lobed, as described by the 
various authors. In some specimens, however, we have 
noticed a five-lobed corolla. Notwithstanding this, these 
flowers had only two calyx-leaves, and the corolla was 
distinctly bilabiate, with the lower lip three-lobed. In such 
flowers there were but four normally developed stamens, with 
a rudiment of a fifth, corresponding to the supernumerary fifth 
corolla-lobe. The fact that some botanists have been led to 
believe that the plant is Orobanchaceous may be due to their 
having observed these abnormal flowers, and also from the 
singular foliage in connexion with the coralloid root-system. 
Another point of interest which so far seems to have been 
overlooked, is that the corolla of Obolaria bears nectaries. 
When the corolla-tube of a living specimen is opened, there 
is distinctly visible at the base of each stamen, almost midway 
between the base of the filament and the base of the corolla- 
tube, a small, lobate, papilliferous scale (Plate XIX, Fig. 1 ). 
As this scale was found in all the flowers we have so far 
examined, it doubtless is a constant character. 
Gilg has given a number of figures illustrating the nectary 
in the genus Sweertia. There is a great similarity between the 
nectary of X. mnlticaulis , Don, and that of Obolaria virginica , 
and therefore Gray is wrong in saying that the corolla of 
Obolaria has no appendages. Baillon’s statement, 4 les fos- 
settes disparaissent dans les Bartonias et les Obolarias,’ is, on 
the other hand, nearly correct, since the grooves in the corolla 
of Obolaria are very imperfect. Bentham and Hooker have 
placed our plant among the Sweertieae, with ‘ corolla efoveo- 
lata ’ ; Gilg has separated our genus from the Sweertieae and 
their relations ; and Knoblauch has followed Bentham and 
Hooker. The indications are, however, that Obolaria is 
closely related to the genus Sweertia. 
Another fact which has not been previously noticed is that 
glandular hairs (PI. XIX, Fig. 6) are present in the sinuses of 
the corolla. Baillon speaks of the presence of such hairs in 
the axils of the bracteoles (calyx-leaves), and Knoblauch 
