i) Cell Union in Herbaceous Grafting. 287 
brown wall noted in the section (fig. 2, /,/). This brown 
wall in all the instances observed was unaffected by stains 
used on the section. As will be seen farther on, this wall 
tends to disappear with age until at last only a mere trace of 
its existence is left. 
In this case the tangential slices, removed in bringing the 
stock and scion into shape for grafting, carried with them a 
part of the woody zone of each internode, so that in cross- 
‘ection the remaining part appears horseshoe shaped. In 
binding the scion to stock the tips of these woody rings were 
bound closely together (fig. 3), but in process of union 
parenchymatous tissue developed and intervened between the 
Woody zones of the two members, resulting in their wide sepa- 
ration, aS seen in figs. 2 and 4. 
In sections of potato grafted to potato no new points were 
noted, the union occurring by the process described in the 
Previous case, except that the brown wall which so strongly 
prenetion in the case of the tomato was not nearly so 
direct} 
_— (Raviti 
len 
wall. In 2 wall twice the thickness of the ordinary cell 
Mentary du €r places this wall was very ragged and frag- 
line by ac to dead cell walls which had been crowded into 
Thegrowsh pone tissues; shown in fig. 5. 
of meristem ofthe cells in closing up gaps is similar to growth 
} Outside the tissue induced by surface wounds. The cells 
| the severed © Path of the knife elongated in the direction of 
thea form * Sela to double the ordinary dimensions or more, 
: ec Stties of transverse walls making what appeared 
me : 
brger pation to be series of narrow plate-like cells, the 
Simension hate were parallel to the cut surface, the longer 
; they origina ‘38 €qual to the width of the cell from which 
the section ae (See figs. 5, II, etc.) “At the outer part of 
© central parenchyma of the stock was shown to 
