a little above the point where these isRUC from the axis ; they arc ovoid, or 

 shortly fusiform, two or three growing together, and each contains a few 

 large, irregularly placed tctraspores. Antlierldia are formed on free pro- 

 cesses of the ramcUi, at uncertain points of the network ; they are oval, con- 

 taining minute granules. The colour is a pale rose-red, soon discharged, 

 and fading in the herbarium to a dirty red-brown. The substance is mem- 

 branous, not gelatinous, and the frond closely adheres to paper in drying. 



Tliougli this plant does not strictly agree in structure with 

 Tlmretia quercifoUa, it accords in so many principal points that 

 I prefer placing it in the genus Tlmretia to founding a new genus 

 for its accommodation. The chief points of contrast are the per- 

 fectly flat penninerved network of T. quercifoUa, and the terete 

 and obsoletely nerved network of the present species, and the 

 shape and relative position of the stichidia. In T. teres the sti- 

 chidia seem to spring directly from the principal axis, instead 

 of from the lateral nerves ; but this can hardly be considered a 

 generic distinction, for it relates merely to the greater or less 

 branching of the axis. No ceramidia have yet been found on 

 T. teres, but some of my specimens produce an abundance of 

 what I suppose to be antheridia, as represented Fig. 5. 



Fig. 1. Thuketia tekes, — the natural size. 2. Cross section of one of the 

 smaller branches of the network. 3. Portion of a fertile costa, bearing stl- 

 cJiidia. 4, Ternate stichidia and base of network. 5. Portion of network 

 bearing antheridia : — variously magnified. 



