FRUITS AND SEEDS 241 



Columbine {Aquilegia), the spurs of which are perforated by 

 humble-bees. The same is doue by them in the flowers of 

 Weigelia, whereas the honey-bee enters the flower. In the 

 different species of Orchis, the spur will often be found ripped 

 open, and in Tropccolum majus there are often two or more 

 holes in the spur, made by these insects. 



Water-plants, too, are fertilised by insects, e.g., Nymphea, 

 Nupliar, Utricularia, Aldrovandia, Hottonia, Stratiotcs, Trcqja. 

 This, however, is only the case when the brilliantly coloured 

 petals are expanded above the water. If the flowering stalks 

 remain short, so that the flowers remain below the water, then 

 eleistogramy is resorted to. In this case small rudimentary 

 hermaphrodite flowers are formed which do not open, and must 

 therefore of needs be self-fertilised. In such cases self-ferti- 

 lisation is generally very effective, and in some cases these 

 cleistogamous flowers are the only fertile ones, while the con- 

 spicuously coloured and open flowers remain sterile. Among 

 land plants, the Violet is the most well-known example of 

 cleistogamy. Of our wild flowers, the Dead Nettle {Lammm), 

 the Eest-harrow (Ononis), the Toad-flax (Linaria), the Wood- 

 sorrel (Oxalis), and others, are provided with cleistogamous 

 flowers, which are not only smaller, but have rudimentary 

 petals devoid of scent and of nectaries. 



Attention must be paid to all these conditions for any 

 success in one of the chief lines of horticultural enterprise, 

 namely, in the development of new varieties by crossing. In 

 spite of many successes, we must admit that the latter are due 

 more to chance than to any carefully planned method. It is 

 true that in this field science has as yet only a few data to go 

 upon and no general principles to lay down, and it might 

 derive much benefit from the observations of horticulturists. 

 In horticultural practice it is generally not possible to take all 

 the precautions of removing the stamens from the flower which 

 is to be fertilised, to protect the flower after fertilisation, and to 

 keep a detailed account of all the cross-fertilisation performed, 

 and very often an assumption of certain processes having 

 taken place has to do duty for actual observation. Of course 

 this leads to mistakes. If we acted quite honestly, we ought 

 often to assert that the origin of certain cultivated forms is 



Q 



