G. R. Wielcuid — Cycadeoidean Flower-bud Structure. 133 



If indeed such a conception or theory of the seed has in it 

 elements of truth, the characterization of a seed as a " mega- 

 synangium," which I would understand to carry with it the 

 idea of a course of direct evolution, rather than the process of 

 reduction just outlined, either fails to express the main fact or 

 is wrong. Only the nucellus can theoretically correspond to 

 the synangial structures. Moreover, if seed coat origins, like 

 floral origins, have largely been processes of reduction which 

 began in Silurian or early Devonian time, the gap between 

 ferns, the evolution of which must conversely be regarded as a 

 more distinctly ascendant process, and the primitive gymno- 

 sperms largely disappears. 



* * 7f * * * 



It does seem then, that there is a certain analogy between a 

 staminate flower and a seed of the older gymnospermous type, 

 that indeed both have a common history as derivatives of 

 homosporous crowns. But necessarily the conception of the 

 course of change which had carried these reproductive struc- 

 tures far apart even in Devonian time can only be cleared by 

 the discovery of fossil evidence. So vast and multitudinous 

 must be the course of change involved that it appears nearly 

 impossible to hypothesize given stages in floral and seed evolu- 

 tion, though certain factors of change do come into view. In 

 the first place, it seems that up to a certain point the process 

 was one of nearly universal forward evolution, after that one 

 of reduction, or of sterilization, whichever one chooses to call 

 it, coupled with branching or budding, — nature taking the 

 course of least resistance, rather than accomplishing the end 

 directly by the continual evolution of new structures. Accord- 

 ing to this new view the actual amount of structural evolution 

 is far less than once seemed could possibly have sufficed. The 

 stony layer of the cycad seed is simply the modified outer 

 sclerenchyma of a series of component rachides, the " blow-off " 

 layer of various forms, the persistent fern chaff or ramentum, 

 and the Lagenostoma cupule, as I have earlier supposed, a 

 bract component. The seed canopy is formed by the original 

 crown of fronds, while the lagenostome even has to do with 

 the prefloration of these. On such an hypothesis of structure 

 the bundles and the ribs are likewise easily explained features, 

 especially if fusion between two whorls be considered to have 

 taken place where there is both an inner and outer flesh tra- 

 versed by bundles. 



It would then seem clear that the first great crest in the suc- 

 cessive waves or pulsations in the evolution of plant, and espe- 

 cially of floral structures, was marked by what could be 

 hypothesized as the Eo-forest, or primitive forest,— the first in 



Am. Jour. Sci.— Fourth Series, Vol. XXXVIII, No. 224.— August, 1914. 

 10 



