G. R. Wieland — Williamsonian Tribe. 441 



oidea stem with a similar though far heavier wood zone, and 

 a very narrow cortex. But in hoth eases the structure, though 

 apparently a sharp departure from the ordinary type of col- 

 lateral woody wedge cylinder almost universal in the cycads 

 hitherto observed, was left in doubt. Now, however, it has 

 been found from the simultaneous examination of thin sections 

 of the Black Hills trunk figured, and of the " coal-oaiV Cor- 

 daitean trunk Mesoxylon (24, 28) that the wood of these forms 

 is strikingly alike. Thus those who have seen Mesoxylon. will 

 recall that though a solid stem, it also has a decided simulation 

 of growth rings which largely disappears under the microscope 

 and is there found to be due to some areal variation in resin 

 content, or to staining during mineralization ; and precisely 

 the same condition is present in the Gycadeoidea stem. 



That is to say, the wood of the Cycadeoidean [and William- 

 sonian] stems was at times, even in the more robust types, not 

 only as heavy as that of Cordaites, but exhibited the same 

 compact structure ; thus permitting the open habitus of plants 

 with small freely branching stems, thin cortex, and finally, 

 excision or non-retention of old leaf bases, as in the conifers. 

 And from this notable approximation of Cycadeoidea to Cor- 

 daites, taken with the known occurrence of small stemmed 

 forms and the even more significant complementary fact that 

 following the seed fern stem types of the Paleozoic there is 

 that long series of Medullosan stems, it may be deemed 

 virtually proven that the Williamsonian stem had, as a rule, a 

 fairly compact wood zone and that it was normally slender and 

 branching, and in the smaller thin-barked armorless forms 

 could very readily be mistaken for coniferous types. At least 

 it may be expected that when the discovery of a more extended 

 series of Williamsonian stems does finally come, solid types of 

 wood, even those with annual rings, will prove abundant, 

 rather than the lax lattice of collateral bundles as developed in 

 the Cycadaceae, and most Cycadeoidece. It is becoming evi- 

 dent enough that instead of being representative of the Cycad 

 vegetation of the past, these robust stems with thin wood, an 

 immense medulla, and heavy persistent armor, are exceptional 

 to the point of abnormality. They do not represent the char- 

 acteristic type of cycad vegetation in the past, which is reason 

 enough for their relatively rare occurrence in ancient rocks. 



Foliage. (Figure 5.) 



Admittedly the greatest bar to progress in the study of 

 Williamsonian forms arises from the extreme difficulty of 

 homologizing the abundant isolated cycadophytean foliage and 

 fruits, as usually associated not only with other fragmentary 



