G. R. Wieland — Cycadeoidean Flower-bud Structure. 129 



With exactly this possibility in view, I have in describing 

 the apical region of the flower as a dome, which it in reality is, 

 had in mind the further thought of analogy to the " canopy " 

 of the seeds of the Paleozoic cycadofilicaleans. Moreover, if 

 one but considers the sterile tips of these fronds in a semi- 

 expanded position they would, if considered as meeting and 

 fusing edge to edge, correspond to the lagenostome of old 

 gymnosperm seeds. This in Lag eno stoma ^ where its most 

 typical form is seen, comes to the surface, rising as high as the 

 elements of the canopy. Continuing, therefore, the analogy 

 already suggested by the use of the terms 'dome' and 

 ' canopy,' for the outer apex of flower and seed, the staminate 

 frond tips taken collectively would be correctly named the 

 cycadostome. If any true homology should ever be estab- 

 lished between these structures these terms might still be used, 

 no better nomenclature suggesting itself, the important point 

 being that sterilized megasporophylls enclosing a megaspore 

 may also be conceived to take the same outer form as the 

 staminate disk. 



For the present, suffice it to say then, the comparisons 

 invited by these striking features are highly suggestive. Any 

 solution or hypothesis of seed origin is of the most excessive 

 difficulty because seeds with the highly developed coats are 

 already present in the Devonian. In fact, the lack of fossil 

 forms giving any definite hint of the mode of seed coat origin 

 beyond the assuredly significant fact that the oldest seeds are 

 the most complex has seemed to make seed coat origins doubly 

 difficult of explanation. And yet the enormous structural 

 gap between the oldest pteridophytes and the earliest gymnos- 

 perms has not deterred investigators from suggesting that seed 

 and pollen-bearing organs are, broadly speaking, mega- and 

 microsynangial structures. With a somewhat different possi- 

 bility in mind, I have already pointed out the fact that the 

 seeds of Neurojpteris are of the most leafy type known, and 

 appear to be in some way related to the microspore-bearing 

 disk Codonotheca, one of the most fundamentally important 

 ancient fruits ever discovered, and one which should be familiar 

 to every paleobotanist. It must be significant that the pollen 

 of Codonotheca is the largest known, but the general fact that 

 ancient pollen is large and shows evidence of prothallial struc- 

 ture was long since sufficiently proven by Renault and also 

 illustrated by Brongniart. 



Is it not, then, fully significant that as we go back into the 

 remoter Paleozoic, seed structures are leafy and have highly 

 developed vascular bundles while likewise pollen is of relatively 

 enormous size % Does not such fossil evidence as is afforded 

 by the seed -like exterior of Codonotheca with the large size of 



