﻿244 RELATIONSHIPS. 



mere half dozen well-understood great plant types scattered over vast periods of 

 time, and representing but a few of a vast array of unknown evolutionary steps, 

 should be exactly the ones enabling us to say, for instance, that certain lines 

 (cycadofilices) led into the Cycadales and Ginkgoales, and sent off a branch which 

 yielded Cycadeoideau stock first, then the Cordaitales, or vice versa, and that from 

 these latter the augiosperms sprang. And no one fact involves the problem more 

 than the unequal rate of evolution in different organs. This has never before been 

 so well illustrated as now. In Cycas, carpophylls of the simplest types known in 

 seed plants persist and are borne on the stem itself. Yet in this genus the micro- 

 sporophylls are reduced to scales and borne on a secondary axis, as are both micro- 

 and megasporophylls in the other genera. Conversely, in the Cycadeoideau it is the 

 microsporophylls that remain primitive while the megasporophylls are organized 

 into a strobilus in many respects far more complex than that seen in any living 

 cycad. Moreover, in this case the microsporophylls, though of such primitive type, 

 have, in assuming the cyclic arrangement so seldom seen in gymnosperms and so 

 often in angiosperms, made possible a wholly new series of reductions. Now, we 

 are certainly permitted to conjecture the former existence of members of a cyca- 

 deoidean alliance in which the microsporophylls did not assume a cyclic arrange- 

 ment and others with perchance freely branching, less compacted trunk types in 

 which there was an early reduction of these spirally arranged microsporophylls to 

 a filamentous staminate form, with, as may often happen, a certain coordination 

 between decrease in bulk and increase in number. It was with this thought in 

 mind that I formerly suggested a certain analogy between the cycadeoideau flower 

 and that of Liriodcndron, which I yet regard as very much to the point. 



Again, if, as we surely may, we assume the former existence of cycadeoidean 

 bisporangiate types in which, coordinate with freer and freer branching and the 

 production of more numerous bisexual flowers, the ovulate cone bore a more and 

 more reduced number of seeds until only a single terminal one was left, another 

 series of highly suggestive relationships presents itself. Moreover, in this final 

 member of such a series bearing one-seeded flowers, a great variety of possibilities 

 would be bound up in the extent to which the micropylar tube elongated and 

 conditioned the process of fertilization and prothallial growth and change, as like- 

 wise in the extent to which reduction aud fusion went in the organs present in the 

 cycadeoideau fruit as interseminal scales. But let us take a step farther. The 

 writer has constantly urged that the great variety of stem types suggested by the 

 freely branching Anomozamitcs minor of Nathorst is a factor of the first impor- 

 tance. Just as the pinnules of a pinnate frond begin to branch and a bipiunate frond 

 arises, so a stem branches and the organs of fructification that were formerly inserted 

 directly come to be borne on secondary and tertiary axes. There is no funda- 

 mental difference between the carpel-bearing summit of the trunk of Cycas and 

 the complementary staminate strobilus or the mega- and microsporangiate strobili 

 of the cycads in general. These variations have arisen in the simplest manner by 

 accelerated branching, coupled earlier or later with sporophyll reduction, when the 

 entire phylum was of a much more primitive type than now. This simple view 

 rids us of many difficulties; and it is very clear that the "plexus" to which the 



