PALEOBOTANY BERRY. 357 



pollen is slightly winged and is formed in microsporangia, which are 

 normally in pairs, pendent from the expanded tips of sporophylls 

 arranged in loose catkins in the axils of scale leaves at the end of 

 short shoots. The mogasporangia are also normally in pairs, although 

 but one ovule matures, and are at the end of a long stalk which is 

 morphologically a shoot and in some Mesozoic species bore more than 

 two ovules. The sperms are ciliated and free-swimming as in the 

 Cycadophytes and Pteridophytes. Ginkgo was formerly considered 

 a member of the family Taxaceae, although its resemblance to 

 cycads has long been known. It appears to have been derived from 

 the Paleozoic plexus of Pteridospermophyta and shows points of con- 

 tact with both the Paleozoic Cordaitales and the Mesozoic William- 

 soniales. All three groups probably had similar ancestors, but there 

 is no evidence that the Ginkgoales have been ancestral to any existing 

 group of Coniferales. 



The modern Ginkgo is the most isolated as well as the oldest exist- 

 ing arborescent form, its stock being represented in the fossil record 

 continuously from the Permian to the present. The leaves are so 

 characteristic that there is little uncertainty regarding their deter- 

 mination, although in some of the older associates referred to the 

 Ginkgoales the foliage simulates such digitate fern fronds as Schi- 

 zaea, and some of these are of questionable identity. The limited 

 space forbids more than a mention of the extinct genera that have 

 been referred to this order and which in some cases do and in other 

 cases do not belong to this stock. Some of these genera are Ginkgo- 

 phyllum, Saportaea, Trichopitys, Dicranophyllum, Khipidopsis, 

 Whittleseya, Psygmophyllum, Gomphostrobus, Trichophyllum, Feil- 

 denia, Phoenicopsis, and Czekanowskia. 



Some of these, such as Saportaea and Khipidopsis, seem clearly 

 to represent early variants of the Ginkgo type, but the two fossil 

 genera that stand out beyond all question are Ginkgo itself and the 

 allied genus Baiera, the latter segregated on account of the short 

 petiole and the repeated dichotomy of the leaf blades to form narrow 

 elongate ultimate segments in which the veins no longer fork and 

 because of the more numerous micro- and megasporangia. The mod- 

 ern Ginkgo sometimes furnishes instances of more than the usual 

 number of pollen sacs and ovules, and its leaves also frequently 

 become divided to simulate those of Baiera. 



Both Baiera and Ginkgo appear as far back as the Permian and 

 become abundant and varied throughout the Triassic, Jurassic, and 

 Lower Cretaceous, where frequently the seeds, immature fruits, and 

 the pollen-bearing catkins, as well as the leaves, are found fossil. 



Baiera attains its maximum range somewhat earlier than Ginkgo, 

 and, as the accompanying map shows, its distribution appears to 



136650°— 20 24 



