IRIDEAE 427 



bee, while smaller insects, especially certain frequently occurring hover-flies (Rhingia 

 rostrata L.), neither bend back the stigmatic lappet nor touch the anthers. Such 

 a visitor, as Hermann Miiller describes, and I have often observed, walks on one 

 of the sepals to the nectar-entrances, inserts its proboscis (11 mm. long), first into 

 one, then into the other, and after drinking, goes some steps backwards, to eat 

 also. As soon as the insect comes beneath the anther, it raises its head, stretches 

 its long proboscis up to it, and devours pollen. Its visits to flowers of this form are 

 thus not only useless, but actually injurious. 



(b) syrphophila Knuth. The stylar branches are situated close to the corre- 

 sponding sepals. Small entrances to the spaces thus enclosed are left below the 

 stigmatic lappets, owing to the arched form of the stylar branches. These are large 

 enough to admit a hover-fly of medium size, and especially the Rhingia men- 

 tioned above, while a humble-bee cannot force its way in. In form (b) a hover-fly 

 behaves exactly like a humble-bee in form (a), creeping under the stylar branch, 

 first brushing against the stigmatic lappet and then the anthers with its back, drinking 

 from both nectar-passages, backing out of the flower without devouring pollen, and then 

 flying to another. Humble-bees are equally injurious visitors for form (b) as Rhingia 

 for form (a). Hermann Miiller observed a large parasitic humble-bee (Psithyrus 

 vestalis Fourcr. $, 25 mm. long and 10 mm. broad) going round the flower 

 repeatedly, inserting its proboscis and sucking laterally above the base of the free 

 part of a sepal into one of the two nectar-passages, so that neither anthers nor 

 stigmas were touched. 



(c) intermedia Knuth. The distance between the sepals and the stylar branches 

 is intermediate between those given for (a) and (3). This form is rare. 



According to Hermann Miiller's observations, humble-bees do not creep out of 

 form (a) backwards, as Sprengel has described, and as I have frequently seen, but 

 after sucking all the nectar from one of the three double tubes in each flower, they 

 shorten the way to another considerably by gripping one of the two adjacent sepals 

 laterally and climbing on to it. They then force their way under the stylar branch, 

 suck nectar, and proceed in the same way with the third nectar receptacle. They 

 next fly to another flower, to behave in the same manner ; thus cross-pollination only 

 is eff'ected in doing this. Hermann Miiller only saw humble-bees back out of flowers 

 of form (c) to settle on another sepal of the same or a different flower. This 

 intermediate form combines the disadvantages of the other two, for it is neither 

 protected from theft of pollen by Rhingia nor from theft of nectar by Bombus. This 

 accounts for its comparative rarity. Warnstorf describes the pollen-grains as yellow 

 in colour, globular when examined in water, with a network of tubercles, very large, 

 on an average 125 /^ in diameter. 



Visitors. — Herm. Miiller (H. M.) and Knuth (Kn.) observed the following. — 



A. Hymenoptera. Apidae : 1. Apis mellifica Z. 5 (Kn., H. M.), vainly looking 

 for nectar, or at most reaching the top of it; then backing out sideways ; 2. Bombus 

 agrorum F. 5 and 5 (Kn., H. M.), skg., observed by Kn. more frequently than 

 the following; 3. B. hortorum Z. 5 and 5 (Kn., H. M.), do.; 4. B. rajellus .AT. 5 

 (H. M.), do. ; 5. Osmia rufa Z. 5 (H. M.), do., but too small to effect pollination ; 

 6. Psithyrus vestalis Fourcr. 5, stealing nectar. B. Diptera. Syrphidae : 7. Rhingia 

 rostrata Z. (Kn., H. M.), skg. and po-dvg. 



