60 
there is but one specimen representing the autumn transforma- 
tion! In V. Manda these runners are ahnost roots, being more 
or less under the surface of the ground, slender, producing no leaves 
and bearing no cleistogamous flowers. They grow from the main 
rootstock, are not numerous, and are somewhat decumbent. V. 
primulicfolia has runners that sometimes reach a length of i3 inches. 
They are comparatively stout, run along the surface of the ground, 
and are mostly leaf- and flower-bearing throughout. These runners 
have nodes or joints from each of which the leaves and erect cleisto- 
gamous flowers start — usually one of each. They root at these nodes, 
and, during the month of October, break away from the parent plant, 
the nodes forming the nuclei of new plants for the next year. In 
nearly every specimen the cleistogamous flowers far outnumber the 
others. The typical form of V. lanceolata has somewhat shorter and 
more robust runners, which are also more leafy— sometimes appear- 
ing almost like incipient branches. While the latter two species are 
invariably found in company, the former is more often solitary, which 
is another point in its claim for specific distinction. 
In all three species the runners begin to form about the middle of 
August, reach their full growth about the beginning of October, and 
then most of them decay except at the points where they have taken 
root. 
Viola odorata produces both runners and cleistogamous flowers. 
The flowers are quite numerous, starting from and clustering around 
the main rootstock. They are very much appressed, sometimes ap- 
pearing as if subterranean. The runners, although having leaves, do 
not seem to bear cleistogamous flowers. 
Viola canina^ var. sylvestris^ shows a very beautiful transformation. 
The branches of the season become elongated, sometimes to full seven 
inches, and bear clusters of cleistogamous flowers, each upon a short, 
slender peduncle in the axils. In some specimens little branchlets 
start from the axils, bearing both leaves and flowers. The autumnal 
peduncles are not more than two inches long, while those of the early 
spring are generally three or four inches in length. Again, in the 
spring, there is but one flower from each axil, while in the autumn 
there are two or more. 
Viola pedata apparently does not produce cleistogamous flowers, 
but It very commonly blossoms a second time in the autumn. Spec- 
imens have been collected as late as the middle of November. 
Mr. Schrenk supplemented the paper with some notes on the struc- 
ture of cleistogamous flowers. 
Mr. Britton referred to a peculiar form of Viola blanda ( V. anmna, 
Leconte) with petioles and peduncles flecked with red, 
Veronica Biixbaumii, Tenore, is reported by Mr. Britton as 
becoming a weed in many places. It is very common in gardens at 
New Dorp, S. L, and specimens have just been received from New 
Brunswick, N. T. 
