344 METAMORPHOSIS 



which develops in the following year all its leaves having been laid down in 

 the autumn. The development which occurs in spring resolves itself essentially 

 into a longitudinal growth, leaves and internodes acquiring their definite 

 length, and this elongation may take place in the case of the oak, beech, &c., in 

 the course of a few days (Josr, 1891 ; KUSTER, 1898). We may describe this 

 as a spasmodic evolution, and it is evident that a certain degree of temperature 

 is needed to give the impulse for the commencing of this development, but that 

 the cessation of this evolution depends not on external but on internal relations. 

 Although new buds obviously begin to form in the axils of the leaves, or termin- 

 ally, by May or June, the leaves of these buds unfold only in the following 

 year; it is impossible, however, for any one to believe that the external conditions 

 are unsuited for their immediate development. We may readily attribute there- 

 fore the non-unfolding of these rudiments to correlation of growth. If a shoot 

 be deprived of its leaves in early summer, in most cases a second evolution of 

 leaves and buds takes place (GoEBEL, 1880), and many trees, such as the horse- 

 chestnut, develop shoots a second time in autumn if they have lost their leaves 

 owing to unfavourable circumstances such as drought. Correlation between 

 the fully-developed leaves and the rudiments of next year's growth, prevents 

 an immediate evolution of the latter. In this case a mere retardation of the 

 functional activity and not complete removal of the leaves is all that is neces- 

 sary to induce the correlation to make itself apparent. Hence one sees that 

 chestnuts or maples grown in darkness (Josr, 1893) proceed to form from the 

 terminal bud, not merely a second, but even a third shoot, and we cannot 

 doubt that by appropriate means, e. g. by accumulation of a sufficient amount 

 of materials, an even greater number of shoots may be induced to form. A second 

 shoot, the so-called ' Lammas -shoot ', certainly appears regularly in many trees, 

 e. g. in young oaks, if the foliage leaves are fully active. In such a case, owing 

 to internal relations not fully known, the correlative inhibition is removed earlier 

 than usual. The characteristic feature, however, of this entire group of plants 

 is that they may be made to increase their production of leaves but never to 

 keep on doing so. Every shoot formation is followed by a period of rest, however 

 short, and the successive shoots are separated from each other by scale-leaves. 

 We have to deal with a peculiar and unmistakable periodicity in the plant which 

 manifests itself not only in the periodic succession of leaf-development but 

 also in quantitative variations in the leaves and internodes. (Qualitative 

 variations we shall refer to later.) Very frequently one sees that the internodes 

 in a yearly increment increase from below upwards and then decrease : 



Length of internodes in cm. 

 Aesculus 3.5 6-5 7.5 0-5 



Acer pseudoplatanus 0-5 4-5 10-5 9-5 7-0 2-5 2-5 o-o 

 Fagus sylvatica 0-4 i-o 2-2 3-2 4-7 7-0 80 8-0 8-0 4-8 4-5 



Similarly the size of the leaves varies (beech) or they may (Acer, Aesculus) 

 decrease regularly from base to apex (compare TAMMES, 1903). 



Another type of shoot evolution is seen in those plants which go on de- 

 veloping leaf after leaf the whole summer through, and in which the cessation of 

 leaf-development is probably induced by external factors. Thus roses grown in 

 a greenhouse go on forming leaves until late in December. This type is natu- 

 rally of less interest ; it need only be noted that it is related by transitional 

 forms with the type previously described. In Forsythia, for example, there are 

 short shoots which have completed their leaf-formation in early summer, and also 

 long shoots which develop leaves right on into autumn and finally die off at the 

 apex without forming any terminal bud. Between these extremes there are 

 shoots which, after a short rest, shoot out anew, and others which after only 

 forming short internodes once more proceed to form long internodes. The deter- 



