i8o INTRODUCTION 



readily visit pollen flowers, since pollen is an important food to them. As a rule 

 they only frequent bee flowers and lepidopterid flowers for the purpose of 

 stealing pollen. 



From Loew's statistical investigations (' Blumenbesuch,' II, p. 120) it appears 

 with regard to the visits to flowers of the more highly specialized flies (exclusive of 

 Syrphidae, Conopidae, and Bombyliidae) that there is a noticeable increase in visits 

 to pollen flowers, bee flowers, and lepidopterid flowers on the one hand, and 

 a diminution in visits to flowers with slightly concealed or exposed nectar on the 

 other hand. 



The Syrphidae and Empidae in the Alps (H. MuUer, ' Alpenblumen,' p. 5 17) devote 

 a not inconsiderable share of their visits to flowers with fully exposed nectar (E), but 

 clearly prefer such as have the nectar partly or completely concealed (EC, C), and 

 affect to a much greater extent social flowers (S), which afford a rich booty. Only 

 a small proportion of their visits are paid to bee flowers, lepidopterid flowers, 

 anemophilous flowers, and pollen flowers. Among the pollen-eating Syrphidae, 

 however, a considerably larger proportion of visits is naturally paid to the latter than 

 in the case of the Empidae, which only suck. Moreover, the preponderance of white 

 and yellow flowers over the red and blue is almost exactly the same in both cases 

 (about 70 : 30 %). 



In the large family of Syrphidae the preference for red, violet, and blue colours 

 increases with the length of the proboscis in the different species. Considering their 

 marked proclivity to pollen-eating, it is natural that greater specialization should lead 

 to increasing preference for social flowers, which yield rich supplies of both pollen 

 and nectar at the same time. Only Rhingia, with a proboscis ii-i2 mm. in length, 

 makes any considerable use of its ability to despoil bee flowers and lepidopterid 

 flowers. — ' Of the families of Diptera besides Syrphidae,' continues Herm. MuUer 

 (' Fertilisation,' p. 4 1 ) in his description of the structure and use of the proboscis ', ' the 

 Muscidae, Stratiomyidae, Bombyliidae, Conopidae, and Empidae are of some impor- 

 tance in the pollination of flowers. The species of the first two families both suck 

 nectar and devour pollen : those of the last three only suck nectar. 



' The pollen-eating Muscidae and the Stratiomyidae have the same soft, cushion- 

 shaped swelling of the end-flaps, and the same armature of chitinous ridges upon 

 them, as the Syrphidae ; and, in spite of some structural differences, they use their 

 mouth-parts for feeding in the same way as hover-flies, and similarly retract them 

 when at rest into a cavity below the head. 



' The species of Bombylius, Empis, and the Conopidae, on the other hand, are 

 purely suctorial. Their end-flaps are not provided with soft cushions beset by 

 homy ridges, but are formed of stiff, chitinous plates, which only serve to guide the 

 suctorial organ, nor can this be retracted into a cavity. 



' We may therefore conclude that the power of withdrawing the proboscis into 

 a cavity below the head is of advantage only as a protection for the pollen-feeding 

 apparatus, and is an indirect adaptation to flower-food, as is the snout-like or 



' A work by E. Becher, ' Znr Keontnis der Mundteile der Dipteren ' (Denkschr. Akad. Wiss., 

 Wien, xiv, 1882), throws doubts on many points in Muller's description (Loew, ' Blumenbesuch,* 

 II, p. no, note 3). I have not been able to see this work. 



