169 



3. With the following plants the anthers remain closed 

 in a space, saturated with water-vapour. In so far as 

 they possess nectaries, thèse latter appeared to exert no 

 influence on the bursting of the anthers. 



Ranunculiis acris L., R. bidbosus L., Aquilegia imlgarîs L., 

 Clematis Vitalba L., Ghelidonmm majus L., Brassica olera- 

 cea L., Géranium molle L., G. Robertianum L., G. macro- 

 rhizum L., Geum urbanum L., Rubus caesius L., Philadel- 

 phus coronarius L., Heracleum Sj^hondylium L., H. lanatum 

 Michx, Aegopodium Podagraria Spr., Carum CarviL., Pim- 

 pinella magna L., Valeriana officinalis L., Ligusirum vul- 

 gare L., Majanthemiim bifoUum De, and Iris Pseudacorus L. 



It is remarkable that Brassica oleracea L. forms an ex- 

 ception to what is otherwise generally observed with the 

 Cruciferae; the position of the stamens with respect to 

 the nectaries which secrète honey abundantly, would make 

 us expect that in a moist chamber they would behave 

 like the others. The same remark holds for the species 

 of Géranium. 



The sécrétion of nectar in the flower attracted the at- 

 tention of various investigators many years before Spr en- 

 gel published his view of the matter. Also after Sprengel, 

 in the flrst half of the preceding century, it has many 

 times been the object of investigation. Ail thèse investi- 

 gators agreed in being convinced that, apart from the 

 signiflcance of the honey-secretion for the fertilisation 

 of the flowers by the intervention of insects, to which 

 Sprengel had drawn attention, the sugar-containing tis- 

 sues and the secreted liquid were still in another respect 

 useful to the plant. 



After Darwin had in 1859 brought to the front again 

 Sprengel's observations on the biological signiflcance of 

 the various properties of the flower — which observations 

 were falling more and more into oblivion — and had ac- 

 cepted their conséquences by bringing them into relation 



