76 HOFMEISTER, ON 



radial to it. In the cells destined to form elaters no 

 transverse septa are formed, whilst their sister-cells by 

 repeated division parallel to the axis of the fruit form rows 

 of cubical cells — the spore-mother- cells. These are, there- 

 fore, arranged in rows, which radiate horizontally from 

 the longitudinal axis of the capsule, each three of which are 

 succeeded by an elater, and of which four (sometimes by 

 displacement, as many as eight), are contiguous to each elater 

 (PL VIII, fig. 7 ; PL IX, fig. 23). 



Contemporaneously with the commencement of the dif- 

 ferentiation of the capsule-wall of /. divaricata from its 

 contents, (which takes place by division into outer and 

 inner ones of three double pairs of cells of the upper 

 clavate portion of the rudimentary fruit,) septa parallel to 

 the axis of the fruit are also formed in the first and second 

 double pairs of cells above the base of the young fruit ; 

 the outer ones of the new cells thus formed divide by 

 radial longitudinal septa. As they expand considerably in 

 breadth, and protrude arcuately outwards, they form the 

 knobby protuberance by which the fruit is attached to 

 the tissue of the stem w r hich bears it (PL VII, fig. 20). 

 The basal-cell and the pair of cells above it remain un- 

 changed. On the other hand, the cells of that portion of 

 the rudimentary fruit which lies between the basal en- 

 largement and the bottom of the capsule, — i. e., the fruit- 

 stalk, — divide all together by horizontal septa. They thus 

 represent quadrants of cylinders of very small height ; the 

 transverse diameter is from six to eight times greater than 

 the altitude. A sudden longitudinal prolongation of these 

 cells, to an extent manifestly far exceeding fifty times 

 the original longitudinal diameter of the cells, lifts the 

 capsule up (when the fruit is ripe), through the fissure of 

 the ruptured calyptra, high above the perianth. 1 have 

 found no trace of a flow of sap into the cells of the fruit- 

 stalk in this species, although it is certainly the one best 

 suited for such an observation. 



The development of the fruit of the greater number 

 of the Jungermannise agrees very closely with that of 

 J. divaricata, and differs only by the more frequent repeti- 

 tion of certain cell- divisions, and by more vigorous develop- 



