Durand : Marchantia polymorpha 323 



definite meristematic regions which lie between the lobes. Goebel 



regards the antheridiophore as representing an entire branching 



system, which divides repeatedly, the apices being near the margin 



and constituting the meristematic regions mentioned. While 



young antheridia may be found on quite large receptacles, it is 



best to collect the smaller ones in which the stalk is still very 



short or even has not begun to elongate. The material from 



which the present study was made was collected May 26, 1900, 



and fixed in Flemming's solution. The sections are 8-10// thick 



and stained in Delafield's hematoxylin. This stain is well adapted 



to work of this kind since it brings out the cell-walls and nuclei 

 very clearly. 



The youngest antheridium that I have been able to recognize 

 definitely as such is shown in figure I. It is a conical cell several 

 cells removed from the margin, and has evidently been cut off by 

 a transverse wall from the cell beneath it. A study of the section 

 shows that the neighboring cells also have been derived in the 

 same manner from the cells beneath. An examination of many 

 preparations of young stages has convinced me that the antheridial 

 rudiment of Marchantia at no time projects above its fellows later 

 to become buried, but is one of the superficial, dorsal, submarginal 

 cells of the receptacle, having the same origin and being cut off at 

 practically the same time as the cells adjacent to it, but soon 

 becoming separated from them, assuming a conical form, and 

 becoming richer in protoplasmic contents. The one shown in 

 figure 1 has already been slightly surpassed by the elongating 

 adjacent cells. 



In the stage shown in figures 2 and j, the rudiment has been 

 divided by a transverse wall into two cells, the proximal being 

 the mother-cell of the stalk, while the distal is the mother-cell of 

 the antheridium proper. Occasionally the receptacle is so convex 

 above that the young antheridia are nearly horizontal so that 

 transections of the gametophore give longisections of the young 

 0r gan (figs, ^6). 



The young antheridium now enlarges (figs. 4, 5) and soon 

 becomes sunk in a cavity owing to the division and upward growth 

 of the surrounding cells (fig. 6). The divisions in the lower cell 

 to form the stalk follow no regular order but may be either longi- 



