865 



Judging by this, Gernet arrives at the conclusion that, even if he did 

 not know the initial stages of the tubers, he must still deny any connection 

 between them and the wood body of the trunk and can consider the tubers 

 to be produced neither from preventitious nor adventitious buds. 



Having investigated the tubers of apple trees, I can confirm absolutely 

 this point of view. For my investigation I had at my disposal tubers vary- 

 ing in size from a millet grain to a pea ; they came from the base of the trunk 

 of a young apple tree, possibly eight years old. The tubers lay in the outer 

 bark, from which they could be easily separated. The tmder side was either 

 completely covered with a smooth bark (Fig. 205, i a) or showed a 

 "brownish, dry point, without any bark and somewhat depressed {i-k) which 

 was surrounded by a green circular bark wall. 



Fig. 205, 2 gives the median cross section of the latter kind of tuber. 

 In this we see a median core {2,h) consisting of two phloem fibre 

 groups separated by a little parenchyma ; other tubers have only one phloem 

 strand in the core, or two or three isolated cores. Around the bundle are 

 deposited cells, parenchymatous in form, with slightly lignified walls and 

 arranged radially. It is evident that they are formed after the manner of 

 cork cells. At times only a group of thick- walled, brown parenchyma cells, 

 with or without starch or phloem fibres, is found in the centre of the tuber; 

 yet thisi is a more rare case. Finally, tubers are formed now and then 

 with a small central cavity, filled with the brown remains of cells. 



The radially arranged, circular zone of lignified, parenchymatous cells 

 passes over gradually into narrow, thick-walled, somewhat elongated wood 

 parenchyma cells, horizontal or diagonal in course, between which lie scat- 

 tered, short, broad vessels with simple pits (Fig. 205, 2,g'). These groups 

 are already divided into numerous circles of vascular bundles by approxi- 

 mately cubical medullary rays deposited in one to three rows. The 

 phenomenon , begins here which continues in alternative zones out to the 

 periphery of the wood body, viz : that the elements of the one part of the 

 bundle, which lies between two medullary rays, show a course differing from 

 that in the adjacent bundle. While the cells and vessels of the one part 

 seem cut crosswise {2 h"), the fibres of the adjacent part are cut longitudin- 

 ally. This is found in trunks which have overgrown some constriction and 

 may be explained only by the theory that the diiTerent parts of the cambium 

 of the wood body, which curves about the core like a shell, are exposed 

 simultaneously to different pressure and strain. Since the young tuber 

 body has no exact spherical form but is only approximately round, the parts 

 which are to overgrow the corners already formed elongate more in the 

 same length of time. 



The elements become narrowed, longer and thicker-walled toward the 

 outside of the tuber until they have nearly the length, form and, in places, 

 arrangement of the normal wood body. 



Inside the tuber, as in the wood, a differentiation of the annual rings 

 into spring and summer wood is found, so that it is evident that the tuber 



