33© METAMORPHOSIS 



the apex of a shoot be removed the bud nearest to the wound proceeds to develop ; 

 if the apex of the root be cut off the nearest lateral root takes its place. As we 

 have already seen, roots may appear on shoots and shoots on roots and even roots 

 and shoots on leaves, and the position of these is governed by definite laws. A 

 branch of willow bearing buds, if kept in a damp atmosphere, produces at its 

 upper end leaf-buds only, and roots only at the lower end, and the size of the 

 shoots increases the nearer they are to the upper end, and of the roots the nearer 

 they are to the base. As Vochting (1878) has shown, the distribution of organs, 

 as well as their relative size, depends on external factors, such as moisture, 

 gravity, and light, though not in the first instance conditioned by these factors. 

 On the other hand, the branch has an inherent polarity of its own, and this may 

 be influenced by external agents but cannot be completely overcome. In Mar- 

 chantia and allied Hepaticae, which exhibit remarkable powers of regeneration 

 (Vochting, 1885), renovation always takes place at the apex of the older part ; the 

 very smallest segments of the thallus show a differentiation into two poles, and we 

 can scarcely doubt that every cell possesses a distinct distal and proximal end. 

 Indeed, Vochting has also demonstrated the polarity of the individual cell in 

 the higher plant (p. 333). We must remember in this relation that in the lower 

 organisms the determination of what shall be base and what apex is frequently 

 settled by external factors — in Bryopsts, for example, by light — but we must 

 admit that the higher plants behave differently. In the phanerogamic embryo 

 and the gemmae of Marchantia the differentiation of base and apex is 

 already established, and depends entirely on internal factors. The phenomena 

 of regeneration teach us, however, that the differentiation into shoot and root 

 taking place after the first cell divisions in the embryo of Phanerogams does 

 not indicate a separation of the protoplasm into two parts, one with root and 

 the other with shoot characteristics, but that every individual cell must have per- 

 manently the capacity of forming both kinds of organ. Only the continuous 

 mutual influence of the parts can condition the formation of a shoot on the one 

 hand or a root on the other. It need scarcely be mentioned that the root 

 behaves in principle just like a shoot, producing shoots at its base and roots at 

 its apex ; further, that the polarity in regenerative processes exhibits itself, 

 very frequently at least, in the leaf, inasmuch as shoots arise on its morpho- 

 logical upper side and roots from its under side. Since this polarity is induced 

 only by correlation, we need not wonder that, in certain cases, it becomes obli- 

 terated where the growing point of the root throws off its cap and produces leaves, 

 in Neottia and Anthurium, for instance, apparently without any external inter- 

 ference, and in Ophioglossum, after separation of the ends of the roots. (For 

 literature see Goebel, 1898-1901, p. 435.) 



The simplest cases of correlative influence are quantitative in their nature, 

 one organ determines to what extent another may develop (' compensation ' ; 

 Goebel, 1884, 1893-5). In the last-mentioned examples, however, we have to 

 deal with qualitative changes, and a few further examples of these may be given 

 here. We may deal first with Goebel's researches on bud-scales (1880). By 

 removing the foliage-leaves at an appropriate time of the year it is possible to 

 inhibit the formation of bud-scales and to transform their rudiments into foliage- 

 leaves, and it is possible to bring about the change of subterranean buds into 

 leafy shoots, as in the potato and many rhizomes, by removal of the main leafy 

 shoot. We shall obtain a better conception of this phenomenon in the next lecture. 



We may next select the spruce as an example of a far-reaching qualitative 

 correlation between the main shoot and lateral buds. The main axis is radial 

 and grows upright, the lateral shoots are dorsiventral and grow obliquely up- 

 wards. It we cut off the main shoot one {or often more) of the lateral shoots 

 highest up the stem grows round as nearly as possible at right angles, and becomes 

 radial in structure. It may be assumed that the dorsiventrality of many lateral 



