364 METAMORPHOSIS 



tion of the problem, since the increase in the intensity of Hght obviously does 

 not play the same part in Vochting's experiments as does the limiting of 

 the food supply in Saprolegnia. A certain light intensity is only one of the 

 conditions of flower development and not the chief releasing stimulus. Under 

 such a light intensity of course vegetative developments go on unchecked. There 

 are many observations and experiments which show that other external factors 

 may be of great importance in the determination of flower formation. Ivy 

 blooms only in a warm sunny aspect, and exhibits vegetative growth only in 

 woodland shade (Wiesner, 1902, p. 75). The quality of the light also has an 

 important effect on flower formation. According to Sachs (1887), ultra-violet 

 light must be especially considered in this relation (p. 311). In his experiments, 

 carried out under conditions when ultra-violet light was excluded, flower-buds 

 were formed, but they did not open ; apparently the intensity of the light was 

 too low. Sachs's statements may be explained by Vochting's experiments and 

 observations (1893) and by those of Klebs (1900 b) and Montemartini (1903). 

 MoBius found (1897) certain Gramineae and species of Borago flowered better 

 on dry soil in places where nutrients were restricted in quantity than where there 

 were abundant supplies of water and nutritive salts. Perhaps with this is 

 related the fact that root-pruning of trees tends to induce an increase in flower 

 formation ; that ringing has also the same effect is probably due in the first 

 instance to the abundant supply of organic material to the axillary buds. 

 [Klebs (1903, and especially 1904, 545) has established the external conditions 

 necessary for flower formation in a number of plants. No general laws on the 

 subject have as yet, however, been formulated ; on the contrary, every plant 

 seems to maintain relationships of its own with the environment.] 



Although external factors are known to be of great importance in flower 

 formation, it is not to be expected that they play as vital a part there as in the 

 lower plants. Even though proof be advanced that flower formation stands 

 in definite relation to a definite external factor, that is not to say that that rela- 

 tion is so simple and so direct as in the case of a unicellular alga. If it be proved, 

 for instance, that a dry soil increases flowering, and that a wet one retards it, we 

 can only say that the character of the soil directly influences the roots and only 

 indirectly the aerial organs. The root may be looked upon as a part of the 

 environment of the shoot, just as every cell is a part of the environment of 

 another cell, but that does not affect the relations existing between different 

 parts of the higher plant, and calculated to affect the formation of flowers also. 

 In fact, the correlations between leaf and flower formation must always be kept 

 prominently in mind. All factors which tend to advance foliage development 

 are unfavourable to flower production and vice versa. In extreme cases flowers 

 are completely suppressed, and certain plants exhibit vegetative growth only ; 

 in certain aquatics Goebel (1893) showed that the luxuriant formation of 

 vegetative organs was the cause of the absence of flowers, and Mobius (1897, 

 p. 137) has brought forward evidence of a similar nature. We need not enter 

 more fully into these correlations since we cannot bring forward anything sub- 

 stantially new in their explanation. It is always important, however, to re- 

 member that in flower formation not merely external but also internal stimuli may 

 play an important part, and some of Sachs's (1892) observations on this question 

 are very suggestive. In May Sachs made cuttings of Begonia in the usual way, 

 and found that the plants springing from such leaf-cuttings gave rise to flowers in 

 the beginning of November, preceded by a luxuriant formation of foliage-leaves. 

 If, however, the leaf-cuttings are taken from a flowering specimen in the end of 

 July, flowers appear on them in the end of September, but few leaves are 

 previously formed. Several other experiments of the same nature are recorded. 

 Thus Goebel (1901) found that leaves of Achimenes haageana in the flowering 

 condition, used as cuttings, proceeded to form flowering shoots at once, while 



