788 



compact as possible ; but at the end of the year a fissure might be 

 discerned round the base of some of the pyramids, especially of those 

 near the top of the cone. The cone had then become twice as large 

 as it was in September. But the rhomboids which terminated the 

 scales, rising in the form of truncated and tuberculated pyramids, had 

 increased much less in the upper part of the cone than in the lower. 

 Accordingly the scales in the upper part, extending a fourth of the 

 way down the axis, were afterwards found to be barren. Moreover, 

 as the rhomboids in the middle and lower part increased, they ex- 

 tended themselves laterally much more than vertically, and there 

 arose thus a remarkable swelling or protuberance in the part of the 

 cone below that, which consisted of the barren scales. In this respect 

 the cone assumed the appearance of that of an Encephalartus, which 

 is figured in Jacquin's ' Fragmenta,' plate 27. Although the barren 

 scales at the upper part of the cone became gradually less close and 

 compact, they continued firmly attached to the axis until the follow- 

 ing midsummer. Had a male plant been present, it appears probable 

 that under these circumstances fertilization would have taken place. 

 Although the Cycadeae are classed as gymnospermous, their ovules, with 

 the exception of the genus Cycas, are so covered and guarded in the 

 earlier stages of their development, that it is difficult to imagine how 

 the pollen can possibly obtain access to them. But, after the barren 

 scales at the top of the cone have begun to separate, a shower of pol- 

 len, falling on it, would easily make its way through the fissures be- 

 tweeen these barren scales, and, going in the direction of the axis, 

 would come into immediate contact with the summits of the ovules, 

 which are all directed towards the axis and placed at a very short dis- 

 tance from it. In considering the mode of impregnation, it is also im- 

 portant to observe, that, whereas the male cone quickly comes to matu- 

 rity, sheds its abundant pollen, and decays, the female cone, being of 

 much slower growth, remains for many weeks in that state, in which 

 the provision here described is made for the admission of the pollen. 

 About July, 1848, all the scales separated from the axis, beginning at 

 the top of the cone. A coloured wax model having been made of a 

 scale with the fruit upon it in the mature state, copies of it accom- 

 pany the before-mentioned models of the cone. The scales were 

 arranged in eight spirals, each spiral consisting of forty scales, and 

 making two revolutions round the axis. The number of drupes con- 

 taining nuts, was probably about 400, two upon each perfect scale. 

 The scales were weighed as they fell from the cone, and their entire 

 amount was 46 tbs. After all the scales had fallen, the axis was found 



