G. R. Wieland — Cycadeoidean Flower '-bud Structure. 119 



homosporous crowns by one and the same, general process of 

 sterilization, fusion of leaves, and reduction, the relation of 

 gymnospermous cones to angiosperrnous flowers will be found 

 a much closer one than has hitherto been conceived possible. 

 Especially too will it be seen that Cycas is in reality special- 

 ized and very closely apposed to other cycads when it is 

 observed that instead of producing numerous compound 

 branches or cones the seeds are simply distributed laterally 

 along flat blades instead of spirally along cylindrical strobilar 

 axes. Any known form which could qualify as an Angio- 

 sperm ancestor must therefore combine some of the characters 

 of Cycas with those of Cycadeoidea, and this in a remote 

 sense the Gnetaleans do. 



****** 



As every one knows who has examined at all attentively 

 the Yale, State University of Iowa, or National Museum col- 

 lections of Cycadeoideans, or the long series of silicified trunks 

 which from first to last has been obtained from Italy and from 

 the Isle of Portland, the number of fruits of larger size pre- 

 served in entirety is very small. The tendency is for the 

 fruits to be broken away ; entirely where the peduncles were 

 of some length, or partly where the fruit, though borne on a 

 shorter peduncle, reaches any considerable size. Moreover, on 

 studying those lesser fruits which appear to be quite wholly 

 enveloped by ramentum, a considerable number prove to be 

 very young ; others, a very large number, are found to be 

 young ovulate cones ; while a far from inconsiderable number 

 of the axes are found to contain cavities lined irregularly by 

 quartz druses, or to have their tissues broken down. Again, 

 there is much difficulty of fruit study in the case of many 

 important trunks, due to that natural dislike that an investi- 

 gator must have, to a certain extent, of injuring a great speci- 

 men by cutting away one or two fruits instead of studying it 

 systematically after sawing through it longitudinally and 

 obtaining the segments most favorable for completer study — 

 an operation requiring time and resources. Also, the number 

 of conserved disks being at best but small in proportion to the 

 ovulate cones, knowledge of the staminate disk variations in 

 form and structure has accumulated relatively slow. 



It thus transpired that, as related in Part YI of these studies, 

 the very reduced size of some of the flower buds of relatively 

 large trunks long remained a rather hidden character notwith- 

 standing the fact that small and young fruits were amongst 

 the first of the subjects considered important and to which 

 much attention was given ; though inasmuch as the general 

 characters of the trunks indicate marked specific variety, it has 

 not been at any time a source of surprise to find considerable 



