1889.] BOTANICAL GAZETTE. 121 



anthers and stigmas. The use of these petals seems to be to 

 protect the pollen from intruders. Sprengel saw bees collect- 

 ing pollen of D. Ajacis, and I have seen a very abundant 

 and useful visitor, Svnhalonia speciosa, collecting pollen of 

 D. tricorne, but I am convinced that they behave improper]} 

 in "so doing. Humble-bees, which are best adapted to fertil- 

 ize the flowers, never gather pollen. On the other hand. I 

 have seen Andrenidae trying to collect it, and they were only 

 hindered by the lateral petals. But for these petals most ot 

 the pollen would be carried away by little bees which would 

 only visit flowers in the male stage. 



The spur of the upper sepal is crumpled, and sometimes 

 fits the spurs of the petals so loosely that its tip is empty and 

 hangs down. Indeed, in one case I found the upper sepal 

 entirely empty, and the spurs of the petals stood in front of 

 its lamina. Delpino (178) regards the spur of the sepal as a 

 protection against the jaws of insects which might attempt to 

 cut a way to the nectar, but both he and Riches 4 found some 

 species of Delphinium to be perforated. 



As in D. elatum, the spurs of the petals are entire at the 

 tips and open below into a common cavity. The nectar, 

 therefore, is held in two receptacles, and I have observed that 

 when Bombus and Synhalonia insert their proboscides into 

 the spur, they regularly draw back a little and thrust then- 

 tongues in again, evidently to extract the nectar trom both 

 petals. I think the double nectary is to favor bees, which are 

 intelligent enough to drain both sides, while butterflies will 

 probably leave one side full. This structure might also be ot 

 advantage in case of perforation, since the robber would have 

 to make Wo holes or leave one side full. The double nectary, 

 however, causes delay, and this seems to be the reason why 

 the nectariferous petals of D. Consolida have developed a 

 common cavity throughout. * 



studied irTbeing male in the first stage and in being specially 



adapted to humble-bees. . . . . ., _ 



The spurs of the petals from the point of insertion to the 

 closed part measure from 7 to 9 mm., and to the tipstiom 15 

 to 17 mm., so that a proboscis 7 to 9 mm. long is neeciea to 

 reach the nectar, and one 15 to 17 mm. long to exhaust it 

 Only the females of Bombus are flying while the plant 

 blooms. B. Pennsvlvanicus, with a tongue 16 to 17 mm. 

 long, is best adapted to suck up all of the honey. 



■•Science Go ip, 1877, 24! 



