Development of the Antheridium in Ferns. 243 



ends (fig. 18). In the smaller of the two sister cells a further 

 division sometimes takes place. It is either divided into two 

 equal parts by a wall perpendicular to the last-formed one, or 

 an oppositely curved wall attaches itself on both sides to the 

 first wall. The opercvdum is then composed of a central and 

 two peripheral cells. Rarely the second wall of the operculum 

 is parallel to that first formed. 



At the opening of the antheridium the operculum is not ir- 

 regularly ruptured as in Aneimia hirta, Ceratojyferis thalic- 

 troides, and Asplemum alatum, but the smaller cell, or, when 

 it consists of three, one of the two smaller cells, is separated 

 from its union with the neighbouring cells, and thrown back 

 like a valve. 



The structure of the ring cells, so far as I could observe, is 

 perfectly analogous to that described in Asj^lenium cdatiim ; 

 here also the lower one is essentially different in its origin 

 from the upper one. The lower one is cut oft' directly by a 

 funnel-shaped septum from the primitive mother cell of the 

 antheridium, whilst the upper one, Avitli the operculum (which 

 is afterwards pluricellular) , is the product of division of a bell- 

 shaped cell. 



The process of development of the antheridia of Osmvnda 

 regalis differs completely from the examples above described. 

 Closed ring cells never occur in it. The mother cell is first of all 

 divided by an oblique wall, which is slightly concave inwards 

 and is followed in the upper and larger of the two sister cells 

 by a second wall inclined in the opposite direction ; only in 

 rare cases three successive walls are formed, and these then 

 diverge at angles of 120°. Whilst the peripheral cells undergo 

 no further division, in the inner and at the same time superior 

 cell a septum, nearly perpendicular to the lougitudinal axis of 

 the antheridium and slightly concave below, is formed, and 

 attaches itself to the first-formed septa on all sides. The 

 central cell is then broken up by a series of divisions, in which 

 no definite rule can be recognized, into the special mother cells 

 of the spermatozoids ; the opercular cell is divided at the same 

 time, by several walls running in the same direction across its 

 vertex, into three or four cells, the outer contour of which 

 usually becomes waved by subsequent extension. They form 

 the greater part of the wall of the antheridium*. 



The interest attaching to the facts above communicated 

 goes far beyond the developmental history of the Ferns. As 

 far as I know, cells in the form of closed rings have only been 

 observed in the mature fronds of some species oi Aneimia^ 



* I shall give a more detailed account of tlie antheridia of Osmmula in 

 a memoir wliich will shortly ajrjiear in Pringsheim's ' Jahrbuch.' 



