FLORAL VARIATIONS AMONG SURREY VIOLETS 147 
and V. Riviniana. I think this well accords with what we know of 
these species. Generally speaking, the first-named species are less 
variable than the other three. JV. odorata is decidedly less variable 
than V. hirta, and, of the allied species, V. silvestris is a more stable 
plant than V. Riviniana. This latter view may not be shared by 
seem to me but variations of V. Riviniana, or of hybrid origin. In 
a certain locality in Surrey such appar — cer mediate forms are 
plentiful, and, if V. hn were comin e of these might 
mistaken for it. Howe , V. silvestris is ea stasis and even 
in tint of colour of sowves is readily recognized by any who are 
acquainted with these species in the field. In the autumn time I 
have found, in the — see rig excellent examples of the hybrid 
V. Riviniana % tied with the cleistogamous flowers totally 
sterile. The difference "be ewe the habitats affected by V. Rivi- 
niana and V. siteiend is may in part account for the fact that variation 
is more frequent in one than in the other. The former species 
a 
heaths, downs, hedge-banks, and copses indifferently ; whilst its 
ally seems more at home on hedge-banks and copses on chalk and 
limestone soils. In the account of the variability of both these 
violets the most abundant scar recorded was the occurrence 
1000 of the 
al- 
more developed in the lower petals, and i opposite the greater 
developed angles of the cohering stamens. 
wk Pipopk: bind few reeebone are to be found among the 
flow f V. odorata may be because vey species is decidedly a less 
freely flowering Heer than Skhoe violets. In the South of England 
may be found in flower, but towards the end of April eas are 
searce. V. hirta closely f wilews the sweet violet in the time of 
opening its first flowers, but here the period of flowering is <a 
longed into June, though these June-time flowers are much dwarfed. 
V. silvestris lags but very slightly behind these two species. Its 
period of flowering is less than that of V. Riviniana, which flowers 
later, and remains in bloom much longer. V. ericetorum is the 
cng “] see violets under sinaiaavation to come into flower. 
nal nefetrtc no doubt pari influence the production 
of sesemtions among violet flow Such conditions are chiefly 
eavaens ms esis Tight, the effect of an unusual 
excess of illumination was well seen in the case of V. silvestris 
growing in a copse which had been cut over during the previous 
winter. In the spring abnormal flowers were abundant. In woods 
L 2 
