FLORAL VARIATIONS AMONG SURREY VIOLETS 141 
combined carpels. The stamens, by their arrangement, form 
a 5-angled body, with the two upper angles smaller than the lower 
three. As the stamens alternate with the petals, it follows that an 
angle will be opposite each petal, which, in consequence, is very 
slightly hollowed - its lower part. The petals are SO —— aa 
two appear to be at the upper part of the flower, one ach s 
the fifth <a peas the lowest of the w Bids is forties 
towards its base with a hollow tubular outgrowth, varying in shape 
and size, termed the spur. Inside the spur are two elongated 
more or less club-shaped or es arising from the short filaments 
of the two lowest stamens. 1e characteristic appearance of the 
usual violet flower is due to the diverse sizes of the petals as well 
as to their shapes and the positions they assume, the result being 
that in one plane only, the vertical, can the flower be divided into 
two parts which reproduce each other 
The number of flowers to be examined of each species was 
fixed at 1000. 
Viola odorata L. 
1000 flowers, gathered on the northern lower slope of the North 
Downs, in the neighbourhood of West Horsley and East Clandon, 
experience, more variations, and of a diverse nature, might have 
been expected. In the preceding — among 200 flowers gathered 
at another part of the North Downs, at Marden Park and n cs 
bourhood, I detected four instances of variation as follows: a), Ww 
each lateral petal bifid at the upper edge ; (4), with all five ats 
shallowly bifid; (c), with lateral petals obscurely 8-lobed and 
lower ‘eal similarly lobed ; (d), with six sepals, six petals, two of 
them, each bearing a spur, appearing to replace the lower petal, 
six stamens, and normal number of carpels. Inside one spur were 
two anther-processes, and inside the other one process. In this 
last flower it will be seen that there is an addition of one member 
to each floral whorl—except the carpels. 
V. hirta, L. 
It was the abundance and diversity “ variations among the 
flowers of this species that induced me take up the subject. 
Ceosiiaislt I had met with flowers ie of the petalline 
flowers, I am convinced, would have been great. * One plant bore 
25 expanded gia s, all of them departing from the usual type. 
eae flower bore two spurs, 15 bore three, one bore four, and eight 
e five. In all these the supernumerary spurs were somewhat 
files than the normal spur, In the case of flowers with three 
