PART III. | THE MECHANISMS OF FLOWERS. ; 443 
its genus. The presumption that this results from self-fertilisation 
is increased by watching the plant when protected froin insects. 
When the flower opens, the anthers are seen to have already de- 
hisced, and to have their pollen-covered surfaces applied to the 
stigma. Ihave myself observed that when insects are excluded 
the plant regularly bears good seed. 
The honey is secreted and sheltered as in V. Chamedrys; the 
stamens are however not thinned at their bases. Even in case of 
insect-visits, cross-fertilisation is not rendered more probable than 
self-fertilisation; but when it does occur, its action probably has 
the mastery. 
I have only seen this flower visited by insects occasionally in 
the first sunny days of spring; later in the season, the competition 
of other flowers causes it to be neglected. 
Visitors : Hymenoptera—A pide : (1) Andrenaparvula, K. 9,8. ; (2) Ha- 
lictus nitidiusculus, K. 2? ; (3) H. leucopus, K. 2; (4) H. albipes, F. 9, all 
three rather abundant, sucking on a small slope where few other plants were 
in flower (April 11, 1869). 
Veronica agrestis, L., possesses the same floral mechanism as 
V. Chamedrys, but in an imperfect or rather in a retrograde 
condition (590, IIL). 
329. VERONICA SERPYLLIFOLIA, L.—The flowers are rendered 
more conspicuous than those of V. hederm@folia, by dark violet lines 
Fie. 151.—Veronica serpyllifolia, L. 
1 —Flower, front view. 
2.—Diagram of ditto : 
a, Stigma; b, stamen; ¢, petal; d, sepal. 
on the superior and lateral petals. They are doubtless more 
visited by insects, and more frequently cross-fertilised ; and accord- 
} _ ingly self-fertilisation does not take place at so early a stage. 
In many flowers, the anthers are at first closed, while the stigma 
is fully ripe; in this condition, insect-visits naturally lead at 
once to cross-fertilisation. -In most flowers, on the other hand, 
the stigma and anthers ripen together; the anthers stand more 
