43 



cellular spaces usually enlarge so as to form small clefts before any 

 cells die : but these clefts lengthen to not more than the diameter of 

 one to three cells before the bounding elements collapse. In the 

 lower internodes where the pith cells all live for a time after secon- 

 dary growth has begun, the cells usually collapse without more separ- 

 ation than that given by intercellular spaces at the angles of the 

 cells. But examples are not wanting in which while the cells remain 

 living the intercellular spaces in these internodes grow into clefts 

 after secondary growth has been going on for some time. This ex- 

 tension of intercellular spaces can probably take place only by the 

 contraction of cell-walls in the neighborhood. Cavity-enlargement can 

 in the internodes where primary growth has ended proceed only by 

 the shrinking and collapse of cells, while in other internodes the 

 movement of all peripheral tissues from the centre carrying with them 

 the enlarging outer pith- cells is an important factor. The difficulty 

 in drawing a boundary between schizogenous and lysigenous origin 

 of cavity that was pointed out in Vicia where the extremes are far 

 apart is much greater in the last-named plants where the extremes 

 are so little removed and where the passage of one form into the 

 other is insensible. 



Urtica dioica and Dahlia variabilis are good examples of the 

 varying relation of cavity- formation to the completion of primary 

 growth. In both of these species the stronger plants will show a 

 cavity much earlier than the weaker ones. The average plant in the 

 former when G or 7 internodes above the earth will show cavity in 

 one of the middle internodes, but weak individuals have been found 

 12 internodes in length with 8 or 9 fully elongated and with solid 

 pith everywhere. A strong Dahlia plant will have its middle inter- 

 node hollow when 3 internodes are exposed above the soil, but weak 

 plants may reach a height of 7 internodes with the lower 3 or 4 fully 

 elongated and all without cavity. Thus in these and doubtless in 

 many other plants, corresponding internodes in different individuals 

 of the same species become hollow sometimes during primary growth, 

 and sometimes not till weeks after secondary growth has begun. 

 The early death of the cells at the centre of the stem would thus 

 seem to be favored by the greater intensity and amount of primary 

 extension in the more peripheral tissue; or in other words, before se- 

 condary growth begins, the less the extension of the peripheral rela- 

 tive to that of the central tissue, the longer will the latter live. 



The experiments in encasing young stems in gypsum have led 

 to the same conclusion. Experience has shown that if the pith is 

 to be preserved much longer than it would live normally, the casts 



