Vol. IV, No. 4.] Notes on the Pollination of Flowers. 201 
(N.S.] 
examining a large number of flowers, it seems to me unwise to 
distinguish these as varieties ; 
] 
i <a they are rather the result of 
Se different conditions. Viola 
canescens 18 the common violet 
ery wher d 
to. very many flowers, and which 
» _ that they serve, instead of the 
narrow lower petal, as the in- 
; have many times observed 
Bombylius major to rest its 
12 feet on them while sucking 
Fig.—9. Lowest petal of Viola Patrinii. from the flower. ae ae 
Figs. 10 and 11 of V. serpens. Fig.12 of thrice I saw Bombylius to rest 
Viola canescens. ‘ on the keel, suck or try to 
ck, and then in each case 
it withdrew its proboscis and revisited the flower, resting “the 
second time on the side petals 
ne of the chief differences between Viola canescens and Viola 

Figs. 13 and 14,—Vigla serpens—flowers in which the upper petals take the 
position usualin V. canescens. x 2. 
none of these differences are absolute guides in diagnosis, but still 
the two species are in a general way distinguishable. The shape 
