PALEOBOTANY BERRY. 351 



divided, giving them a graceful fernlike appearance. This is ex- 

 hibited in Stangeria in another way. Each pinnule has a thick 

 midrib with several vascular bundles, and the broad lamina with 

 prominently toothed margins is traversed by dichotomously branch- 

 ing lateral veins running to the margins, which give it such a fern- 

 like appearance that the first specimens of the foliage to fall in the 

 hands of botanists were described as a new species of fern of the 

 genus Lomaria. In Cycas the narrow pinnules have a midrib with 

 a single large vascular bundle. All the other genera lack a midrib 

 and have the pinnules traversed by subequal longitudinal bundles. 

 The order of development and the venation of the fronds vary from 

 genus to genus. 



The leaves are in whorls crowning the stem or the branches in 

 branched forms, giving a strikingly handsome appearance. Each 

 whorl alternates with a whorl of scale leaves. The whorl may con- 

 sist of but a few to over 100 fronds, varying in size from a length of 

 10 centimeters to more than 3 meters and the fronds may have but 

 a few to as many as 250 pinnules. The structure is xerophytic and 

 the pinnules generally have entire margins. They vary from lance- 

 olate to linear and acicular. In some Zamias and Stangeria the 

 margins are serrate. In Dion they are thorny, and in Eneephalartos 

 they are inequilaterally lacerate. 



A feature reminiscent of ancient habits and shared only by Ginkgo 

 among existing seed plants is the fertilization of the egg cells by 

 multiciliated swimming sperm (ells. 



The histological and morphological details of the Cycadales be- 

 long to modern rather than paleobotany and need not be dwelt upon 

 in the present discussion. 



Both the staminate and ovulate sporophylls are always aggregated 

 in strobili, which are compact and rather uniform throughout the 

 order, except that m Cycas the ovulate strobilus instead of being 

 compacted in a cone consists of rosettes of sporophylls resembling 

 reduced foliage leaves, in which ovules replace the lower pinnae or 

 teeth. 



The fossil record of the Cycadales is extremely meager. It includes 

 certain not conclusively determined ovulate cones from the Keuper 

 of Switzerland, resembling those of Zamia, and more convincing car- 

 pellary leaves, like those of the modern Cycas, from the top of the 

 French Triassic (Cycadospadix hennocquei), from the Kimeridgian 

 of Italy and Scotland (Cycadospadix pasiantanus), and from the 

 Rhaetic of Sweden (Cycadospadix integer), which demonstrate the 

 presence of true Cycadales in the late Triassic and Jurassic. The 

 Cretaceous records are few and not especially trustworthy. During 

 the Tertiary a few undoubted remains of cycads occur at scattered 

 localities, demonstrating a more extended geographical range than 



