PART III. | THE MECHANISMS OF FLOWERS. 431 
plants fairly fertile, in this case insects-visits by promoting 
_ hybridisation do harm. 
; According to Delpino, the Verbascums are adapted for pollen- 
collecting bees, which, clinging to the hairs on the stamens, clear 
i the pollen off the anthers and fly rapidly to another flower. I 
agree with this view, and look upon it as the most plausible 
_ explanation of the hairy stamens. But from my own observations 
_ it isclear that the small pollen-feeding bees (Prosopis) and Syrphide 
| are very frequently the fertilisers of this genus. Delpino goes too 
far in saying that the flowers of Verbascum are adapted solely for 
bees (esclusivamente melittofili), and that the visits of all other insects 
are accidental and without significance (178, IL, pp. 296-298). 
Calceolaria pinnata, L.— Each of the two anthers is modified, as 
in Salvia pratensis, into a lever, of which one arm bears a_ pollen- 
less anther-lobe and stands in the mouth of the corolla. When this 
- arm is touched by an insect, it causes the other, the pollen-bearing 
lobe, to shed its pollen. Self-fertilisation may take place in the 
falling off of the corolla (352). 
i’ 320. LINARIA VULGARIS, Mill—The flowers which I have 
examined do not agree in all points with Sprengel’s description. 
_ Honey is secreted by the base of the ovary, which is especially 
prominent anteriorly, opposite the lower lip. As a rule the honey 
does not, as Sprengel thought, leave the tip of the spur empty, 
_ flowing down at intervals in large drops which cannot reach the 
' bottom for the air contained there ; but it glides ina smooth, narrow 
groove, bordered by short, stiff hairs, which passes from the nectary 
between the two anterior stamens, and thence to the tip of the spur, 
which it fills to a depth of 5 or 6 mm, or even more. In several 
- hundred flowers which I examined, I found two which corresponded 
in this point with Sprengel’s description, so it seems probable that 
_ he based his account upon an abnormal specimen. 
The adjacent sides of the two inferior stamens are closely beset 
with pointed processes at their base, which protect the honey from 
 Sshort-lipped insects, when such succeed, as ants frequently do, in 
_ entering the flower. The hairs bordering the groove protect the 
_ honey in like manner from insects, and also help to keep it in its 
course ; the length of the spur would be of no advantage unless 
the honey were strictly confined to it. 
The length of the spur (10 to 13 mm.) excludes short-lipped 
__ bees from the honey, and flies and beetles are prevented from 
_ entering the flower by the tumid lower lip, which completely 
