1 84 BOTANICAL GAZETTE [September 



mass than the slender ones above. The uppermost cell arising in 

 the epibasal row divides to form the capsule ; the middle originates 

 the seta, probably by its intercalary growth as Leitgeb has sug- 

 gested. When, therefore, the seta consists of three or four tiers of 

 cells, the capsule is definitely differentiated. It consists of two 

 rows, each of eight cells. Periclinal walls have cut out a wall layer 

 one cell thick (fig. 38), leaving a central sporogenous tissue of eight 

 cells. The lower four divide by horizontal and vertical walls; the 

 upper also divide, but form only a group of sterile cells — a cap, later 

 continuous with the elaterophore. The wall of the lower half 

 becomes two layers by periclinal divisions not at all simultaneous 

 (figs. 39, 40). The difference in rate of growth from now on is a 

 striking feature of development in the capsule (Le Clerc du 

 Sablon 18). It first shows in the contrast between the lower and 

 peripheral region and the upper central part. The contents of the 

 cells differ in size of nuclei and amount of cytoplasm. Cell divisions 

 are in every direction. The capsule changes from spherical to oval 

 and elongates rapidly (fig. 41). This difference in rate of growth 

 accompanies the formation of the elaterophore, but what determines 

 the rate? The more slowly dividing cells of the upper central 

 region begin to elongate, and the elaterophore is outlined (fig. 4 2 ) 5 

 its cells have smaller nuclei and less cytoplasm. Although cell 

 divisions are fewer along its margin, they must still be considered 

 sporogenous tissue. Diagonal divisions and radial arrangement of 

 diamond-shaped cells indicate elongation of the capsule. Cells 

 continuing the axis of the capsule between the elaterophore and the 

 base are still rectangular, like those of the elaterophore except in 

 size of nuclei. 



Differentiation among the spindle-shaped cells is the next 

 evidence of separation of the sporogenous tissue into elaters and 

 spore mother cells (figs. 43-50). While in the central axis the 

 elongated diamond-shaped cells appear to form continuous rows to 

 the base, in the radial peripheral regions this is more uncertain. 

 The next stage, and a most unsatisfactory one for study, shows a 

 partial transformation of the walls of the elaters and of the spore 

 mother cells. The elaterophore forms a central cylinder of long 

 prosenchymatous cells, the marginal ones of which have a free tip. 



