56 THE FERTILISATION OF FLOWERS. [PART II. 
abdominal brushes have adapted themselves to the flowers which 
were fitted to dust their ventral surfaces (Papilionacew, Composite, 
Echium, etc.), and the contrary view, that these flowers have 
become adapted to the bees, is untenable, for the flowers are 
visited and fertilised. by other and far more numerous insects ; 
still more untenable is Delpino’s idea (e.g. in regard to Heriades 
truncorum and Helianthus) of mutual predestination. 
While a pollen-collecting apparatus, as we have seen, has been 
developed in very different ways in the two main divisions of the 
family, the adaptation of the mouth-parts for deeply-placed honey 
has advanced similarly in both. In both, a natural limit to the 
length of the cardines and of the mentum (by which, in Prosopis, 
Sphecodes, Halictus, and Andrena, increased protrusibility of the 
tongue was attained) has been fixed by the length of the head, 
underneath which these parts must be withdrawn to give play to 
the mandibles; and access to more deeply-placed honey is got by 
lengthening the tongue itself, and by extension of the membranous 
and elastic parts between the mentum and the cardines. 
In the higher forms of both groups, we find that the tongue, 
which was at first much shorter than the mentum, and retractile 
into it, is many times as long as the mentum; the transverse 
markings (absent. in Prosopis, slight in Sphecodes, well-marked in 
Halictus) are present as strongly-marked transverse rings over the 
greater part of the vermiform tongue ; the hairs of the tongue, which 
have scarcely a definite arrangement in the lower forms, form whorls 
upon each transverse ring, and can be erected or depressed forwards, 
and the originally membranous and elastic parts between the 
mentum and cardines are elongated and supported by chitinous 
pieces, in such a way that when these fold together the mentum 
is retracted between the stipites as far as the ends of the cardines, 
and can be again protracted for the full length of the chitinous 
pieces. 
With these modifications of the lower lip, certain changes in 
the maxille are closely associated, and advance similarly in both 
divisions of the family. As soon as the tongue has so far increased 
in length that it can no longer be quite withdrawn into the anterior 
hollow of the mentum, it comes to be folded up downwards and 
backwards, and both in the retracted and in the extended state it 
is concealed between the laminz to avoid injury in nest-building 
or in being introduced into honey-receptacles. The laminz, once. 
devoted to the service of the tongue, elongate part passw with the 
tongue itself; and correspondingly the labial palps and at first also 
