Holm. — On Obolaria virginica. 371 
extant, and it is with the hope of adding to the knowledge 
of this plant and of encouraging botanists to find additional 
facts in regard to it that we venture to give here some data 
concerning it. 
Various descriptions of the inflorescence of Obolaria vir- 
ginica have been given. Gray first described it as centrifugal, 
but later described it as a raceme, with the flowers ‘ in clusters 
of three, terminal and axillary or solitary ’ ; Bentham and 
Hooker considered it a spike. However, the fact is that the 
flowers are arranged in three, or, by abortion, one-flowered 
cymes. Often small specimens show only a single flower, 
which is terminal, and at the sides of which rudiments of two 
lateral flowers are usually found. In large specimens the 
number of normally developed flowers may reach twenty-one, 
with rudiments of twelve others, thus corresponding to one 
terminal and ten lateral three-flowered, or, by abortion, 
one-flowered cymes. This is a cymose inflorescence, which, 
according to Eichler, is common in the Gentianeae. 
Usually the flowers are sessile, and the peduncles of the 
inflorescence short, excepting in the case of the lower ones. On 
examining the flowers two free calyx-leaves will be seen, and 
this singular fact has given rise to the supposition that Obolaria 
has no calyx, but two bracts instead. Linnaeus considered 
them a pair of bracts, while Elliott stated that they probably 
perform the function of a calyx. Even Baillon in his mono- 
graph of the Gentianeae described the flowers of Obolaria as 
‘ flores asepali.’ It is true that the calyx of Obolaria is very 
different from that of the ordinary Gentianeae, not only on 
account of the suppression of the two leaves, but also because 
those which are developed are, with the exception that they 
are smaller, similar to the stem-leaves and the bracts, having 
the same spatulate shape, with rounded apex, and exactly the 
same nervation. Their lateral margins are slightly involute, 
forming an envelope around the flowers, thus performing the 
true function of an ordinary calyx. Owing to the fact that 
the position of the calyx-leaves in the one-flowered inflo- 
rescence corresponds to that of true prophylla, being situated 
