RHIZOCARPEyE 



27 



conical point becomes cut off by a septum, forming the neck-canal-cell. 

 Below this a second very small portion of the central cell is again cut off 

 to form the ventral canal-cell^ so that the canal now consists of two 

 cells. These two cells become transformed into mucilage, which escapes 

 by forcing apart the four apical or stigmafic cells, leaving an open canal 

 In the meantime the protoplasm of the large basal portion of the central 

 cell has become transformed by contraction into the oosphere. The 

 archegone is now ready for impregnation, the antherozoids reaching the 

 oosphere through a funnel-shaped depression in the epispore and the 

 open canal. After fertilisation the canal again closes up by the expan- 

 sion of the stigmatic cells. The 

 archegone of Azolla resembles that 

 of Salvinia in all essential points. 



The male prothallium of Salvinia 

 is reduced to a mere rudiment. The 

 microspores lie imbedded in a mass 

 of hardened, granular, frothy mucilage, 

 formed by the disorganisation of the 

 tapetal cells. They do not escape 

 from this mucilage, but the endospore 

 of each develops into a tubular fila- 

 ment which pierces through the muci- 

 lage and the wall of the sporange. 

 The extremity of this filament which 

 projects outside the sporange is 

 curved, and becomes cut off by a 

 septum. The lower and larger of the 

 two cells thus formed is regarded as 

 a rudimentary prothallium ; the termi- 

 nal cell, which again divides into two, 



as a rudimentary antherid. The protoplasm of each of the two 

 antheridial cells divides into four, and each of these eight masses of 

 protoplasm escapes as an antherozoid. Each antherozoid is a corkscrew- 

 like coil of protoplasm, bearing vibratile cilia of great length at its 

 broader extremity. To the same extremity is attached a vesicle, com- 

 posed of a portion of the protoplasm of the antheridial cells which 

 was not used up in the formation of the antherozoids, and which does 

 not leave the antherozoid during the period of its 'swarming.'" 



The development of the multicellular embryo from the fertilised 

 oosperm has been very carefully followed out in Salvinia. The first 

 segmentation is by a nearly vertical wall (at right angles to the surface 

 of the prothallium) into two somewhat unequal portions, each of which 



Fig. II. — Salvin'a nutans. A, micro- 

 sporange, with microspore-tubes st. 

 (x 100.) B, microspore-tube (x 200) 

 with closed, C with empty antherid. 

 D, antherozoids ( x 6co). (After 

 Pringsheim.) 



