364 THE AMERICAN NATURALIST. [Vor. XXXIV. 
of the spreading dogbane. These are a little larger than 77. 
letus, and probably found the entrance way to the flower too 
close quarters. H. detus visits the flowers for pollen only. 
Its proboscis is too short to reach the nectar. It wanders in 
at the entrance, rubbing first the stigma and then the anther 
with its back, swinging its proboscis from side to side, the ter- 
minal flaps sweeping the fallen pollen from the floor of the 
passageway. The down of its back was generally found well . 
dusted with Iris pollen, and as an agent in pollination it was 
certainly second only to the bees above mentioned. 
In the method of their operations these bees and syrphus 
flies stand in striking contrast. The bees enter and leave the 
flowers on the run, visiting very many in a remarkably short 
time; the syrphids loiter about the entrance, turning this way 
and that and, although entering cautiously, remain inside some 
time; but between visits they spend much time disporting 
themselves with their fellows in the sunshine. While a syr- 
phus fly is visiting one flower a bee will visit a score. 
Bombus separatus Cr. —'The Iris and the bumblebee were 
the subjects of one of Sprengel's early studies on the relation 
between flowers and insects. Sprengel thought the bumble- 
bee the only insect concerned with the fertilization of the 
European species (/ris germanica) which he studied. Müller 
attributed rather more importance to a very long-tongued, nec- 
tar-eating syrphid, Rhyngia rostrata. American writers on 
[ris versicolor have treated the bumblebee as the visitor of 
first importance. My observations do not corroborate this 
opinion. I saw but three bumblebees on Iris flowers during 
the entire season, and each of these was seen but a few sec- 
onds, entering from one to three flowers at one side only, and 
then departing. The visits were not such as bumblebees make 
to flowers which they habitually seek. They showed evident 
dissatisfaction. Indeed, they seemed to me ill-adapted. They 
are too large; they enter with difficulty and some delay, and 
crowd the stigma too roughly in passing. In the single visit 
which I was able to observe in detail the stigmatic flap was 
torn from the style for half its width; the bumblebee de- 
scended the passage with its proboscis retracted, made no 
