CYCADOPHYTA 603 



with structure preserved would be likely to throw some 

 light on the question. Unfortunately our accurate 

 knowledge of the structure of these plants in Mesozoic 

 times, though of extraordinary interest and value, is 

 still limited to a comparatively narrow group^a special 

 family which had attained a leading position in those 

 days, somewhat as the Compositae have in the recent 

 Flora. The Williamsonias diverge widely, in habit, 

 from the usual Bennettitean type, but in their case we 

 have as yet no knowledge of the internal structure. 



The discovery of the bisexual nature of the flower 

 and the arrangement of its parts has so greatly 

 strengthened the Angiospermous analogies of the 

 Bennettiteae, that an actual, though not a direct 

 relationship, now appears probable. This important 

 subject, however, will be left for discussion in the con- 

 cluding chapter. 



As regards affinities in other directions, the relation 

 to the Cycadaceae is universally recognised, and has 

 been sufficiently dwelt on above. Further research will 

 be necessary before we can hope to determine the 

 common ancestry of the two lines, with which other 

 Cycadophytic phyla, as yet unknown, may have 

 converged. 



Advanced as the Bennettiteae were in their floral 

 organisation, there can be no doubt that they were 

 still Gymnosperms as regards their mode of fertilisa- 

 tion, as shown by the fact that the micropyles were 

 exposed, protruding through the openings left between 

 the coherent scales of the pericarp. 



The stamens, while in their arrangement and 

 monadelphous fusion they suggest the andrcecium of 



