é PART 111. | THE MECHANISMS OF FLOWERS. 131 
58. LYCHNIS VESPERTINA, Sibth. (L. dioica, 8, L.).—The structure 
of the flower has been thoroughly explained by Sprengel. Honey 
is secreted by the lower, fleshy part of the (developed or abortive) 
ovary, and lies at a depth of 20 to 25 mm. in fertile or pistillate 
flowers, and of 15 to 18 mm. in barren or staminate ones. Both 
_ kinds of flowers are much constricted near their upper end by the 
_ calyx, and can only be forced open a little at the very brim by an 
} insect’s head ; a thin proboscis at least 15 to 20 mm. long is needed 
to reach the honey. The pure white flowers, which have no 
guiding marks, expand in the evening, though they are not 
completely shut by day. These characters taken together exclude 
diurnal insects except such as come seeking pollen, and make the 
- flowers so much the more attractive to crepuscular and nocturnal 
Lepidoptera. According to Delpino (178, p. 161), the anthers ripen 
two by two, a fact which I have not observed. A further remark of 
Delpino’s, that the flowers stand horizontally, and let the stamens 
emerge at their highest part, does not at all hold good of the 
specimens in my neighbourhood, which have an almost perpen- 
dicular position, as Sprengel figures them. 
I have repeatedly seen Sphinx porcellus, L., visiting and cross- 
fertilising the flower. 
59. Lycunis Girmado, L., is, like Dianthus, adapted for fer- 
tilisation by butterflies, and resembles Dianthus in regard to its 
-nectaries, the position of its honey, its Se sean dichogamy, 
‘and the order of development of its anthers (702, p. 254). 
_ Visitors : A. Lepidoptera—Rhopalocera : (1) Hesperia silvanus, Esp., s. ; (2) 
Pieris brassicee, L., s., very ab. B. Diptera—Syrphide : (3) Rhingia rostrata, L., 
I saw it insert its proboscis repeatedly, seeking in vain for honey. See also 
No. 590, IL. 
_ The species of Silene and Lychnis have a relation, according to 
Dr. Buchanan White, to species of Dianthwcia (Noctwide) like that 
ppaich exists between ‘Yucca and the Yucca-moth. The species of 
Dianthecia fertilise Silene and Lychnis, and thereby make provision 
for their larvae, which feed exclusively on the unripe seeds of these 
plants. But Silene and Lychnis have many other insect-fertilisers 
besides (773). 
Orv. CARYOPHYLLEZ; b. ALSINE. 
60. CERASTIUM ARVENSE, L.—The flowers resemble those of 
_ &S. Holostea in the position of their nectaries, in the order of develop- 
ment of the two staminal whorls and the stigmas, and hence also 
K 2 
