864 



In this we have to mention first Gemet's^ investigations of tuber formation 

 in Sorhus aucuparia. He found the dead tubers so loosely attached to the 

 bark that they could easily be lifted out v^ith the finger nail while the living 

 ones were apparently firmly fixed in the sapwood. Nevertheless, they 

 proved to be "completely separated from it and appeared as bodies possibly 

 belonging in some way to the phloem because the very reddish color of their 

 smooth under end corresponds to that of the phloem." Most tubers, when 

 cut through, show several centres about which complete wood layers have 

 developed in 13 to 15 annual layers, provided with vessels and medullary 

 rays and agreeing in their cell structure with the wood of the trunk. The 

 course of the wood layers was gnarly. The annual rings were almost always 

 broader in the under half of the tubers toward the trunk than in the upper 

 one, projecting from the trunk. It was not possible to prove any connection 

 with a bud. Even when a tuber lay near a wen, no connection could be 

 found with any of the many bud cones of the wen. 



Unfortunately, Gernet had no opportunity to study the initial stages of 

 tuber development ; the youngest stages in his material were tubercles 0.5 

 mm. in size, still completely enclosed in the bark, without having caused any 

 external protuberance. They lay outside the phloem fibre and were spher- 

 ical or ellipsoid and showed several centres about which the wood body had 

 already been deposited. This consisted of parenchymatously formed cells 

 in which a differentiation of medullary ray cells became recognizable in 

 longitudinal sections. The first indications of vessels may be considered to 

 be represented by a few cells with large lumina but still lying above each 

 other with almost horizontal, unbroken walls and containing less starch, or 

 none at all. The farther all these cells lay from the centre, the more clearly 

 noticeable became the lessening of their radii and the lengthening of their 

 tangential axes ; their cross section approximated that of summer wood. In 

 older tubercles are found at first sharply differentiated a few pitted vessels 

 and a clearly recognizable central parenchymatous centre, rich in starch. 

 The wood body was surrounded by a cambial zone and its own bark. In 

 the upper half of the tubers, cork formation took place at times in the inner 

 bark. The outer side of this newly produced cork zone was united, not 

 infrequently, with the cork zone of the trunk. The part of the bark isolated 

 by such a cork zone (Gernet's "cork dam") loses its starch grains, becomes 

 filled with air and dies gradually so that the outer side of the tuber body 

 contains dead tissue. As a rule, the appearance of these cork layers also 

 introduces the death of the tuber, which occurs within the next few years. 

 The under half of such diseased tubers, as well as that of perfectly healthy 

 ones, retains its living bark tissue and the formation of the bark body pro- 

 gresses with that of the wood body. From this we may conclude that the 

 tuber grows downward and thus its upper part gradually projects above the 

 surface of the bark of the trunk by rupturing it. 



1 Gernet, C. v., tJber die Rindenknollon von Sorbus aucuparia. Moskau 1860. 



