238 The Botanical Gazette. [August, 
The names of these species have been purposely omitted in 
this paper, and yet it is possible that certain of the species 
mentioned under the third series (III) may have had such an 
origin. This may, for instance, be true of Viola arborescens, 
Ceratonia Siliqua, Glinus lotoides, Erica multiflora, An- 
drosace pyrenaica, Daphne Gnidium, and Triglochin laxi- 
florum : ti 
tember and October. This is remarkably retarded develop- 
‘ment for the flowers, considering that the subtending leaves 
are already formed in spring. On the other hand, these in- 
florescences may be branched, or even be developed together 
with a few leaves on short lateral branches; and no signs ofa 
former existence as a spring flowering tree are shown by any 
remnants of organs protecting these inflorescences from the 
cold. Species of Erica often commence flowering very early 
in southern countries. Perhaps E. multiflora is only an at- 
