Durand : Marchantia polymorpha 331 



enous tissue have separated completely from one another, and 

 have elongated. The great majority are narrowly triangular in 

 outline, and are frequently arranged in pairs end to end. A few, 

 however, are narrow and slender, and present the first indications 

 of a differentiation between sporogenous cells and elaters. In the 

 much larger embryo shown in figure ioi } the distinction .between 

 fertile cells and elaters is more marked. The former have lost to 

 some extent the regular triangular outline, while the latter are 

 much more slender and sometimes flexuous. The two kinds of 

 cells alternate irregularly. 



In its further development the sporogonium elongates rapidly 

 until it becomes oblong or elliptical in general outline. The stalk, 

 however, remains for a long time very slight. At the same time 

 the separate cells filling the capsule increase in size. The elaters 

 become long and fusiform, and the contents arranged in a spiral 

 manner next the wall. While the sporogenous cells increase in 

 size they do not seem to increase in number after the separation 

 shown in figure ioo. They soon divide, however, by transverse, 

 or by transverse and longitudinal walls into groups of eight, rarely 

 four, cells [fig. 102 a-k) 9 which do not separate but remain con- 

 nected together. These are the spore-mother-cells. If the divi- 

 sion is transverse only, the mother-cells are arranged in rows (fig. 

 IQ 2 a t d y e). If longitudinal division also occurs, the mother-cells 

 are biseriate or subbiseriate in the group (b } cj, g, h). A very 

 characteristic condition is shown in figures c and //, in which the 

 original cells were subtriangular in outline and arranged in pairs, 

 a $ appears so often in figures ioo and 101. The resulting groups 

 of mother-cells are also subtriangular in form and still remain in 

 P ai rs. The mother-cells are irregular in shape, and are flattened 

 uhere contiguous ones come in contact. 



The mother-cells now increase in size as the capsule expands. 

 The details of their division have not yet been worked out, but 

 each gives rise to four tetrahedral spores (fig. 103 i-m). The 

 s Pores remain connected in the tetrads for a long time, and the 

 lat ter, as well, cling together nearly as long in their original 

 groups. The tetrads, however, seem to become free from one 

 another before the spores separate (fig. 103 m). 



About the time of the division of the spore-mother-cells, ac- 



