4 Watson, The Cone of Bothrodendron mundum. 



Taken in the light of the great tapering shewn in 

 WilHamson's longitudinal section, these figures suggest 

 that the cone tapers rapidly. 



It was probably about 8 millimetres in diameter at 

 its widest place when uncrushed, and tapered to 3 or 4 

 millimetres at the top. 



There is no good evidence as to its extreme length, 

 but Williamson's longitudinal section, which includes 

 the tip, shews macrospores at its base and microspores 

 in the superior sporangia : it is thus probably a nearly 

 entire specimen, and its length is 10 mm. approx. The 

 cone was thus of very small dimensions when compared 

 with the majority of the Lycopod cones of the Coal 

 Measures. 



St met II re of the axis. 



The wood is hollow and surrounds a pith of ordinary 

 soft parenchymatous cells ; this is as a rule badly pre- 

 served and is of no great interest. The wood is of com- 

 paratively large size when compared with that of an 

 ordinary Lepidostrobjis (See Fig. i of Plate). It consists of 

 an inner layer one tracheid wide of large tracheids, this is 

 succeeded by a belt of much smaller tracheids. 



This belt exhibits a tendency to be split up into 

 distinct groups, each of which encloses a protoxylem. 

 The protoxylem points do not project, but form a quite 

 smooth surface to the wood. 



The whole wood very closely resembles that of a 

 small twig of Lepidodeiidrou {= BotJirodendroii) inundum, 

 Will. 



In one cone on section A 188 a small cjuantity of 

 secondary wood has been developed in an irregulur 

 manner ; this is seen in Fig. 2 of Plate. 



Secondary wood is not yet known in stems of Bothro- 



