COLEOPTERA 187 
violaceus creeping about the inflorescences of an Umbellifer (Aegopodium Podagraria), 
obviously pursuing other visitors. 
Considering the small importance of beetles with regard to pollination, continues 
Herm. Miiller (op. cit., p. 34), it is scarcely worth while to compare all the antho- 
philous species, genera, and families with their nearest non-anthophilous relatives 
in order to discover any possible adaptations to flowers. It will be sufficient to 
investigate one family—the Cerambyctdae—with this object. 
One of the chief groups of this family, the Zef/urddae, includes our native 
genera Rhamnusium, Rhagium, Toxotus, Pachyta, Strangalia, Leptura, and Gram- 
moptera. The large majority of species belonging to these genera are exclusively 
anthophilous when adult, except those of Rhamnusium, which are never seen upon 
flowers, but only on willows and poplars. The species of Rhagium are found chiefly 
on fallen wood, but now and then on flowers; those of Toxotus mostly occur on 
flowers, more rarely on shrubs; and the species of the four remaining genera 
confine themselves entirely to flowers. Parz passu with increasing predilection for 
flower-food are developed those peculiarities in bodily structure which distinguish 
Lepturidae from other Cerambycidae, and which enable them to feed on exposed 
or more deeply seated nectar. The characters in question are: elongation of the 
head forward ; a neck-like constriction behind the eyes, giving the power of directing 
* the mouth to the front; elongation and anterior narrowing of the prothorax; and 
the development upon the lobes of the maxilla. of hairs used to lick up nectar 
(Fig. 76). 
- All these characters present a complete series of adaptational stages, from those 
Cerambycidae which never visit flowers, and those which can only lick tolerably 
exposed nectar, up to Strangalia attenuata, which is able to extract nectar from 
the bottom of the corolla-tubes of Knautia arvensis that are 4-6 mm. long. 
In concluding his account Herm. Miiller remarks that, ‘although the 
Coleoptera are of little importance for the pollination of our native flowers, they © 
are nevertheless of special interest. ‘This’ is because they show very clearly the 
first beginnings of insect anthophily, and the early adaptations correlated therewith. 
We see that among the most diverse coleopterous families, of which the members 
_ vary greatly as to their diet, individual species have first partly and then entirely 
accustomed themselves to flower-food, with the result that variations favourable to 
this habit have been naturally selected.. Transition to the anthophilous habit must 
in some cases have taken place long ago, in other cases more recently, for we 
find on the one hand that sufficient time has elapsed for the evolution of antho- 
philous genera and families—by adaptational divergence—while on the other hand 
we find anthophilous species side by side with sister species that have no taste 
for flower-food.’ 
These details of coleopterous structure as described and luminously expounded 
by Hermann Miller, apply only to our indigenous species. As he himself remarks 
(‘ Fertilisation,’ p. 433, note), ‘Some tropical and sub-tropical beetles present much 
more thorough adaptation to flower-food. Thus, in a species of Nemognatha, which 
my brother Fritz Miiller observed sucking flowers of Convolvulus at Itajaky [in 
South Brazil], the outer maxillary lobes (galeae) are modified into sharp grooved 
bristles (12 mm. long), which when apposed form a suctorial tube like the proboscis 
