68 



shaped cushion, on which are inserted a large number of slender 

 rods or pedicels, each terminating at its apex in a single erect seed. 

 Among the seed pedicels a still larger number of sterile organs are 

 borne, the whole forming a closely packed mass. The sterile 

 organs spread out at the top and coalesce, forming a wall to the 

 fruit, interrupted only by minute apertures into each of which the 

 tip of a seed is tightly fitted. The whole is enclosed by a number 

 of overlapping bracts which spring from the fruit stalk, so that 

 the fruit has the appearance of a large bud. 



The cavity of the seed is filled by a large, dicotyledonous 

 embryo. It is only in the genus Bennettites that the embryo 

 has been found preserved in a fossil state. Its presence shows 

 that the fruit was mature at the time of fossilisation. 



The structure of the flower, as distinguished from the fruit, 

 was first made known by Wieland's researches on the abundant 

 American specimens. He showed that the arrangement of the 

 parts was essentially the same as in a modern bi-sexual flower; 

 on the exterior the overlapping bracts represented the perianth ; 

 next came a ring of large, branched stamens, bearing very 

 numerous compound pollen-sacs, while the middle of the flower 

 was occupied by the ovule-bearing organ, representing the pistil. 

 In detail, however, the parts of the flower were very different from 

 those of any living flowering plant, the stamens, for example, 

 rather resembling the fertile fronds of a fern. 



The nearest analogue among recent flowers is to be found in 

 Magnolia and its allies, but the resemblance is a distant one. 



That family of Cycadophytes, of which Bennettites is the 

 type, were enormously abundant in Secondary times, but they did 

 not stand alone. 



There is also the extensive group of the Williamsonias, dis- 

 covered much earlier; Zamia (now W illiamsonia) gigas, from the 

 Lower Oolite of the Yorkshire Cliffs, having been described by 

 Lindley and Hutton in 1835. Williamsonias are now known from 

 India, Mexico and Sweden, as well as England, and were 

 no doubt of world-wide distribution. They were closely allied to 

 the Bennettites family, but, unlike the latter, appear to have had 

 tall stems. The great flowers, which were probably in many cases 

 uni-sexual, somewhat resembled in appearance the flower-heads of 

 an artichoke. 



In a Swedish Rhsetic fossil, Wielandiella, of this family, the 

 stem was slender and forked, quite unlike a typical Cycad. 



We are led to the conclusion, on the whole of the evidence, 

 that in the Secondary period the whole world was overspread by 

 an immense class of plants, having much in common with the 

 little surviving family of Cycadaceae, but in the structure of their 

 flowers far more highly organised, and somewhat approaching the 

 higher Flowering Plants. In geological age they overlapped the 

 latter, for Angiosperms occur, side by side with Bennettites, in 

 the Lower Greensand and even earlier. 



