CARYOPHYLLEAE 197 



a small amount of nectar seems to be secreted, for insect visitors busy themselves 

 persistently in this region. During my frequent visits to Helgoland in 1895 and 

 1897, however, I was unable to find any nectar, though I examined the bases of the 

 petals under a tolerably high magnification. The sepals, which are beset with 

 glandular hairs, are almost as long as the petals, and help a little in the attraction 

 of insects. 



The flowers are homogamous. At first the anthers project about 1 mm. beyond 

 the stigmas, but near the end of anthesis incline towards and dust them with pollen. 

 This self-pollination is clearly effective, for though insect-visits are exceedingly rare 

 all the flowers set fruits. I always found both stamens and carpels developed, and 

 examined numerous specimens without finding a single case where either kind of 

 sexual organ was entirely aborted. Pentamery predominated in the more vigorous 

 plants, tetramery in the weaker ones. 



An insect visitor thrusts its head into the base of the flower, and in doing 

 so touches the anthers which in bright weather rest on the reflexed petals and 

 also the simultaneously developed stigmas, which are at the same level. Hence 

 a visit to a second blossom must effect cross-pollination. In dull weather the flowers 

 close, so that the anthers come into direct contact with the stigmas, and automatically 

 self-pollinate them. 



In Helgoland, on June 5, 1895, I saw a hover-fly (Syritta pipiens Z.) visiting 

 the flowers. This observation interested me very greatly, for on the morning of 

 the same day I noticed on the high ground of that island the same species (also 

 Eristalis sp.) sucking nectar in the flowers of Cochlearia danica. This confirms 

 to a certain degree the ideas to which W. J. Behrens gave expression in 1878 

 (Flora, Marburg, New Series, xxxvi, 1878, pp. 225-32), with regard to the oecological 

 connexion between Cerastium tetrandrum and Cochlearia danica. The flower 

 mechanisms of the two species agree to such an extent that Behrens thought it 

 possible that the former might have arisen under the influence of an insular environ- 

 ment from Cerastium semidecandrum, mimicking Cochlearia danica, a form better 

 adapted to attract insects. The agreement as to insect visitors appears to give to 

 this supposition a higher degree of probability, although Syritta pipiens also visits 

 other flowers of similar structure in Helgoland (e.g. Brassica nigra and Capsella 

 Bursa-pastoris). This, however, appears natural, as the size of this insect is adapted 

 to these other flowers quite as well as to Cerastium tetrandrum and Cochlearia danica. 



Visitors. Vide supra. 



489. C. glomeratum Thuill. Henslow states that the flowers of this species 

 are autogamous ; they sometimes remain closed (Warming). According to Ludwig 

 (Bot. Centralbl., Cassel, iii, 1880, p. 102 1), there are female flowers, distributed gyno- 

 dioeciously, as well as hermaphrodite ones. Now and then the petals are vestigial 

 (Kirchner). According to Warnstorf (Schr. natw. Ver., Wernigerode, xi, 1896), the 

 homogamous and autogamous flowers open only a little, or remain closed, and the 

 introrse anthers lie upon the branches of the style. The pollen-grains are white, 

 and roundish-dodecahedral, with six well-marked germinating processes around the 

 equatorial zone, and a diameter of about 37 /*. 



Visitors. Schletterer observed Halictus calceatus Scop, at Pola. 



