J PART III. | ‘THE MECHANISMS OF FLOWERS. 161 
_ their anthers now covered with pollen come to lie above the ovary 
in the middle of the flower; as they wither they bend back again, 
_ and two others take their shins: The stigma only ripens after all 
_ the stamens have withered, so that all possibility of self-fertilisation 
is excluded. 
___ My own observations, made upon plants flowering in my room, 
differ from Sprengel’s in several points. In all the flowers which 
I examined, the stamens rose up not in pairs but singly, and 
each was joined by its successor a short time before it withered 
and fell back ; so that when two were erect at the same time, one 
of them was at the close the other at the commencement of its 
_ period. After all the stamens have bent back the stigma ripens, 
but before it withers the stamens rise up again; and if, through 
lack of insect-visitors, they still retain a stock of pollen, they shed 
“some of it upon the stigma. So Treviranus’ assertion that the 
approach of the stamens to the stigma takes place for the purpose 
(3 of self-fertilisation is not quite groundless; but such self-fertilisa- 
tion takes place only when the flowering time is passing away 
_ without insect-visits having taken place. 
The whole arrangement of the flower greatly resembles that of 
_ Parnassia. In both we have fully exposed honey and successive 
_ development of the stamens and then of the stigma; in both the 
_ conveyance of pollen is insured by the centre of the flower, where. 
_ an insect is most likely to alight, being occupied successively by a 
_ pollen-covered anther and a mature stigma; both are visited by 
_ many Diptera and Hymenoptera. But while the white flowers of 
_ Parnassia are visited by beetles, the yellowish flowers of Ruta are 
_ altogether avoided by them. 
_ Visitors: A. Diptera—(a) Stratiomyide: (1) Sargus cuprarius, L. ; (d) 
_ Syrphide : (2) Syrphus ribesii, L., ab. ; (3) S. nitidicollis, Mgn. ; (4) Meli- 
 threptus pictus, Mgn. ; (5) Ascia podagrica, F. ; (6) Eristalis sepulcralis, L. ; 
(7) Helophilus floreus, L. ; (8) Syritta pipiens, L., very ab.,—all mainly 
sucking, but sometimes also eating pollen; similarly (¢c) Muscide: (9) 
 Lucilia silvarum, Mgn. ; (10) L. cornicina, F. ; (11) Sarcophaga carnaria, L. ; 
| (12) S. hemarrhoa, Mgn. ; (13) S. albiceps, Mgn. ; (14) Calliphora erythro- 
eephals, Mgn. ; (15) Pollenia rudis, Mgn. ; (16) Sepsis; (17) Anthomyia 
| radicum, L. » very ab. ; (18) A. obelisca, Men. ; ; (19) A. pratensis, Mgn., (the 
i last three species were determined for me by Herr ‘ecg B. Hymen- 
. 0 ptera—(a) Evaniade : (20) Feenus affectator, F.; (21) F. jaculator, F. ; 
(b) Ichneumonidae : (22) various species ; (c) Chryside: (23) Chrysis ignita, 
_ L.; (d) Sphegide : (24) Crabro oo v. d.L. 9; (25) Oxybelus bellus, 
Dib. ; (26) Trypoxylon figulus, L. ; (27) Rhopalum clavipes, L. ; (28) Agenia 
a pret, F. $; (29) Tiphia minuta, v.d. L. ¢; (e) Vespide : (30) Odynerus 
M 
