138 Transactions of the. Society. 



We have to do then, in this case, with an organ in the stage of a. ^floirer, 

 as distinguished from the/7'ui^ previously described. The ovulifer- 

 ous cone, or gyn<ecium, is completely surrounded by the hypogynous 

 staminate disk, as Dr. Wi eland calls it, springing from the rim of the 

 receptacle at the base of the cone (see diagram, plate IX. fig. 5). The 

 stamens are numerous (18-20 ia C. dacotcnsis) and arranged in a. 

 whorl ; their stalks are united to form a continuous sheath, which 

 extends to about the level of the top of the gynrecium. Here they 

 become free from each other; each stamen is a compound, pinnate 

 sporophyll, about 10 cm. long altogether, and is folded inwards 

 towards the gyUcBciura, the deflexed tip reaching down nearly to 

 its base. The alternate pinnie, of which there are about 20 pairs, 

 are likewise bent inwards. The pinnae, with the exception of 

 those at the apex and base of the frond, which are sterile, bear the 

 pollen-sacs in two rows, 10 in each row on the longest pinnae. 

 Thus the stamens are highly complex structui-es, resembling the 

 fertile fronds of a fern, rather than the stamens to which we are 

 accustomed in our modern Flowering Plants. The complexity, how- 

 ever, does not end here, for each pollen-sac is itself a compound 

 structure containing two rows of loculi, 10 or more in each row. 

 It thus constitutes a synoMfjiiim, comparable to that of the Marattia- 

 ceous ferns, and especially the genus Marattia. The similarity to 

 the fructification of such a species as Marattia Kauljitssii is, in fact, 

 surprisingly close. It may be added that the vascular system of 

 the stamens is a complex one, consisting of numerous bundles, as 

 in the leaf of a highly organised fern or a Cycad. 



It appearsthat all the specimens actually investigated were in the 

 bud condition, the stamens being still infolded, as described al)o\e. 

 Plate VIII. fig. 4 is from one of Dr. Wieland's sections, showing 

 portions of the compound infolded stamens ; from the comparison 

 of a number of such sections the author was able to build up the 

 whole construction of the flower. He presumes that the stamens 

 eventually opened out, and the diagrams introduced in figs, o and 6 

 show them in the expanded condition. The ground-plan of the open 

 flower, shown in plate IX. fig. 6, is based on Gijcadeoidea ingens, a 

 species in wliich a number of stamens is smaller than in C. daco- 

 tcnsis. 



It is an interesting fact that in a large number of fructifications 

 in the more mature, seed-bearing condition, Dr. Wieland was able 

 to detect the remains of the staminate whorl seated on the rim of 

 the receptacle. It thus appears that the hermaphrodite condition 

 was general, if not universal, throughout the group, a conclusion 

 which explains Count Solms-Laubach's early observation on 

 Cycadcoidea ctrusca. As the frait matured the expanding gynajcium 

 evidently encroached on, and ultimately filled, the space originally 

 occupied by the staminate whorl (compare plate VII. fig. 2, A, with 

 plate IX. tig. 5). Dr. Wieland is of opinion that the flowers must 

 have been protandrous, the pollen being shed before the ovules 



