96 



osis of the wood, ruptured on the outsid 

 ing years. The edges of the over- 

 growth, still connected in the first few 

 years, have grown back farther and ' 

 farther; — in this way, a crater-like 

 opening was produced at the top of the 

 woody tuber. The new annual rings 

 turn to resin every year and always in 

 the first spring wood, which consists in 

 part of parenchymatically forme 1 

 cells. The resin holes (H) are pro- 

 duced by the drying up of the resini- 

 fied tissues, in part also by exudation 

 of the resin. The edges of the over- 

 growth are further apart each time so 

 that the last ones (U) are widely sep- 

 arated. In this, they show a most ir- 

 regular construction often changing 

 between every two medullary rays in 

 the same annual ring. In the drawing 

 G is the normal wood in cross-sec- 

 tion and M the regular course of 

 the tracheids in longitudinal section. 

 These are in the same annual ring just 

 as in true gnarls. 



I'^or this reason these structures 

 musi be classed with the resin galls. 

 So far as their production is concerned. 



e and was overgrown in tlie follow- 



Fig. 8. Crcis.s-section through a resin gall on the 

 stilt-like root of the pine. (Orig.) 



Fig. 7. Resin galls with gnarl 

 growth on the upper side of the stilt- 

 like root of the pine (natural size). 

 (Orig.) 



t must be assumed that the exposed 

 root shows small centres of 

 injury from extremes of 

 weather on its upper side, 

 i. e. the one most exposed 

 to such extremes. These 

 centres of injury have 

 caused a resinosis of the 

 tissues, or rather, a com- 

 I)lete resinous liquefaction. 

 We may assume that frost 

 has caused the injuries, and 

 especially late frosts, since 

 these appearances are al- 

 ways found in the first 

 formed spring wood. The 

 production of these resin 

 galls shows that the roots 



