FLORIN: SYSTEMATICS OF THE GYMNOSPERMS 337 



situated on either side of this fold. The microspores of Taxus and Juniperus 

 represent a reduced form. 



Burlingame (1915a; cf. Goebel, 1932; Schnarf, 1933) distinguished four 

 methods of pollination and fertilization in gymnosperms : 



1. In the extinct Cycadofllices and cordaites the pollen grains lodged in a pollen 

 chamber in the nucellus, and probably did not develop pollen tubes. 



2. In the cycads and Ginkgo they lodge in a pollen chamber in the nucellus, and form 

 haustorial branching pollen tubes, which do not penetrate towards the female gametophyte 

 or take any part in transferring the ciliated sperms to the archegonia. 



3. In most of the conifers the pollen passes down to the tip of the nucellus, where it 

 puts out a pollen tube as a sperm carrier. 



4. The araucarians are pollinated on the ovuliferous scale at a distance from the ovule, 

 from which point a pollen tube grows towards the micropylar end of the ovule and there 

 enters the protruding nucellus. 



Coulter and Chamberlain in 1917 gave an account of the rather meager 

 knowledge of gymnosperm embryogeny at that time. The real advance in this 

 field is mainly due to Buchholz (1929), who in 1918 began publishing a series 

 of contributions relating to the conifers. Considering mainly the early stages 

 between the proembryo and the organization of the embryo into tissue regions 

 and organs, Buchholz realized that many of the diverse features of conifer em- 

 bryogeny are variations due to tlie complications of cleavage (monozygotic) 

 polyembryony, a condition occurring, together with simple (polyzygotic) poly- 

 embryony, in this group, in contradistinction to that of the cycads, where only 

 the latter type of polyembryony is found. He regarded as primitive features 

 an extended period of apical cell growth in the embryo, cleavage polyembryony, 

 and rosette embryos. The absence of apical cells, simple embryogeny, the ab- 

 sence of rosette cells, and the presence of embryonic caps, were believed to rep- 

 resent the most advanced features of conifer embryogeny. 



Pteridospermae 



One of the most important events in the history of paleobotany was the first 

 recognition of the seed of a member of the paleozoic Cycadofllices. This seed, 

 Lagenostoma Lomaxi, was enclosed in a cupule bearing capitate glands iden- 

 tical in form and structure with the glands on the associated vegetative organs 

 of Lyginopteris {Lyginodendron) oldhamia (Oliver and Scott, 1904). The seed, 

 of complex structure, is orthotropous, radially symmetrical, and borne termi- 

 nally on the ultimate and naked ramifications of the frond. The group name 

 Pteridospermae was introduced, and later (Scott, 1923) proposed for exclusive 

 use in place of the name Cycadofllices. In 1904 White announced the terminal 

 attachment of small seeds to pedicel-like pinnules of the frond of Aneimites fer- 

 tilis, and in 1905 Grand' Eury published his find of similar seeds on the frond 

 of Pecopteris Pluckeneti, where they occurred at the margins of almost un- 

 modified pinnules. Kidston (1904) found large seeds attached terminally to a 

 pinna rachis of Neuropteris heteropJiylla. In 1911, Kidston and Jongmans de- 

 scribed the even larger seed of Neuropteris hollandica. Much indirect evidence 

 was brought to light, from association and comparative structure, in support of 

 the assumption that such frond genera as Neuropteris, Alethopteris, Lonchop- 

 teris, and Linopteris represented pteridosperms with medullosean stems, and 



