602 STUDIES IN FOSSIL BOTANY 



carpels, but there seems to be no reason for assuming 

 this, and the transitional forms between scales and 

 bracts are against it. On the view here provisionally- 

 adopted, the whole lateral axis constituting the flower 

 may be compared to the main axis of a Cycas, or rather 

 to that of a hypothetical Cycadophyte or Pteridosperm 

 bearing both male and female sporophylls, the leaves 

 adjacent to the sporophylls being reduced, as are the 

 scale- leaves of Cycas. The bracts are but little 

 modified, while the stamens fully retain their foliar 

 characters, and have been well compared by Wie- 

 land to the megasporophylls of Cycas. The inter- 

 seminal scales and the monospermic carpels have 

 undergone a somewhat extreme reduction, and their 

 arrangement appears very different from that of the 

 whorled stamens, though a verticillate order may 

 possibly yet be traced in the organs of the gynaecium. 

 The reduction demanded by the carpellary theory is, of 

 course, far less than that which the axial theory of the 

 pedicel requires ; on the latter view each monospermic 

 pedicel would be the homologue, not merely of a 

 carpel, but of an entire strobilus. 



As compared with the seed of a Pteridosperm or 

 Cycad, that of the Bennettiteae is considerably simpli- 

 fied, especially in its vascular system, which appears to 

 be almost limited to the chalazal disc or socket. This 

 reduction is no doubt correlated with the enclosure of 

 the seeds in a fruit. 



Our knowledge of the earlier Cycadophyta is at 

 present quite insufficient to enable us to form any 

 conception of the course of evolution of the Bennet- 

 titean flower. The discovery of Triassic Cycadophytes 



