Studies of Selected Spermatophytes. 241 



about the drops of nectar excreted by the nectary. Thus 

 it attracts the drops from many sides and consequently 

 more powerfully than the nectary does, and the drops 

 of nectar must follow the stronger force and flow from 

 the nectary into the end of the spur. And here from the 

 same cause the nectar must remain and not flow down and 

 out of the corolla as it tends to do on account of its weight. 

 . . . That the nectar is fully protected from all injury 

 by the rain is apparent. Even if a drop of rain should 

 get near to the opening of the spur, it could not enter ; but 

 in order that it may not even approach the opening the 

 two middle petals have hairs just where they can be 

 most effective. Thus when raindrops have fallen upon 

 the upper petal and, running down it, have united into a 

 single drop, the latter is arrested as soon as it has reached 

 these hairs. It is therefore utterly impossible for a rain- 

 drop ever to reach the nectar. ... In Fig. 13 [in 

 Sprengel's book] one sees the greater part of the nectar 

 guides on the lowest petal. This figure and Fig. 8 show 

 how the veins run a bit into the opening of the spur. 

 Therefore a bee were as stupid as a fly if it did not know 

 how to find the nectar as soon as it has alighted on the 

 flower. Now how is this violet pollinated? In order 

 properly to answer this question, which for several years 

 was an apparently insolvable riddle for me, I must make 

 the reader somewhat more closely acquainted with this 

 flower. The five anthers surround the pistil and conceal 

 it so that one sees no more than the reflexed end of the 

 style. They are not grown together but touch one another 

 and appear to be a single body. The filaments are some- 

 what fleshy ; the two lowest have each a projection just as 

 fleshy which extends into the spur ; the ends of the pro- 

 jections, as said, secrete the nectar. Each filament has an 



