580 THE FERTILISATION OF FLOWERS. [PART LY. 
concealment of the honey the number of short-lipped visitors diminishes and 
the number of long-tongued visitors increases ; and that, in the end, bees far 
outnumber beetles, wasps, and short-lipped flies. 
The tubes in which various flowers conceal their honey vary in 
length between very wide limits—from scarcely 1 mm. to about 
30 mm. in our native flowers. In the shortest forms, ¢.g. species of 
Veronica, the tube with its interior circle of hairs acts just like 
the structures which shelter the honey in Malva and Geranium. 
The further increase of length in such tubes depends entirely 
upon the advantage to the flower of excluding certain groups 
of insects in order to induce more diligent visits on the part 
of others. 
If increase in length of the proboscis proceeded pari passu with 
increased need of food and increased diligence, and therefore also 
with increased usefulness in the work of fertilisation, the various 
adaptations connected with it would be comparatively easy to 
understand. But this is by no means the case. The entomophilous 
insects which require most food and are most diligent, and there- 
fore the most efficient fertilisers, are unquestionably the Bees, 
since they feed not only themselves but their young solely on 
substances derived from flowers; but the circumstance that they 
have to use their mouth-parts to build the cells for their young 
has set definite limits to the adaptation of these parts for obtaining 
deeply placed honey, while such limits do not exist for Lepidoptera. 
Accordingly, though bees have far surpassed flies in the length 
of their tongue (our longest-tongued Diptera, Bombylius and 
fthingia, have proboscides 10 to 12 mm. long, while Anthophora 
prlipes and Bombus hortorum have proboscides over 20 mm. long), 
they are far outstripped by many Lepidoptera (Sphinx ligustri, 
37 to 42 mm., S. Convolvuli, 65 to 80 mm.). 
Though, owing to the special diligence of bees, most tubular 
flowers have elongated their tubes so as to exclude the more 
short-lipped insects and to admit a wider or narrower circle of 
the more specialised bees, and though many of these flowers 
are shown by other structural adaptations to be specially fitted 
for bees, yet butterflies have access to the honey of all these 
flowers, and sometimes act as subsidiary fertilising agents though 
often they are quite useless to the plant. 
Most of our native flowers with hidden honey have at the same time their 
pollen more or less deeply situated, or else differ in other important relations, 
and are therefore not fitted to show clearly the effect of increased concealment 
8. +? => eae 
