KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 63. N:0 8. 



59 



ser ved; then, the auxiliary cell cuts off a placental cell, giving rise to the sporogenous 

 lobes — all acc. to Lewis, Life hist. of G. Bornetiana. Kylin doubts the correctness of 

 Lewis' observations in this case, and I find it more probable that the placental cell of 

 Lewis corresponds to the auxiliary cell of Kylin and the writer. The carpospores of 

 G. antarctica are formed as in the other species, not as in Bornetia. 



The involucral rays are 5 to 8 (rarely more) in number and formed by the first 



Fig. 28. Griffithsia antarctica: 



—C tops of male plants with spermatangial branchlets in various stages; d lobe of male 

 branchlet, e top of tetrasporic branch. All X 120. 



(basal) central cell as in G. corallina. In G. antarctica they are unicellular, not two-celled 

 as in the other. 



The tetrasporic branchlets are formed round the apex of a subapical 

 cell, encircling the node (fig. 27 c); the terminal cell is abruptly narrowed at the base 

 and does not undergo any further development (comp. fig. 28 e). In the formation of 

 sporangia G. antarctica agress with G. corallina and Bornetiana; there is a stalk cell pro- 

 ducing numerous sporangia, so numerous that already on early stages it is rather diffi- 

 cult to obtain a good view of their arrangement. Fig. 29 k shows three stalk cells with 

 sporangia in various degress of development, fig. 29 1 one stalk cell, where all mature 

 sporangia ha ve been removed. The involucre, of 12 to 15 unicellular rays, originates as 



