FLORIN: SYSTEMATICS OF THE GYMNOSPERMS 339 



to gametophytes, fertilization, and embryogeny, but also as regards the external 

 morphology and anatomy in general of their reproductive and vegetative or- 

 gans. Chamberlain, (Joebel (1923), and Pilger (1926) traced a gradual reduc- 

 tion from Cycas revoluta to Zamia, in the size, form, and branching of the 

 megasporophylls, in the number of ovules, and in the number of microsporangia 

 on each microsporophyll. Stopes (1904) thought that the cycad ovules had a 

 double integument, but Quisumbing (1925) showed that tliis was an error due 

 to the failure to study the stony layer of the testa from its inception. Accord- 

 ing to Kershaw (1912) the closest parallel to the cycads among fossil seeds is 

 to be found in those of Trigonocarpus affinity. 



The bulk of the secondary xylem in cycad stems consists as a rule of tra- 

 cheids with multiseriate bordered pits, but in Stangeria and Zamia of scalari- 

 form tracheids. According to Sifton (1920, 1922) and Bailey (1925) the alter- 

 nate and opposite pitting is directly developed from scalariform types. As 

 regards the general course and organization of the leaf traces in the cortex, 

 Dioon was studied by Langdon (1920) and others. Of the several strands sep- 

 arating from the vascular cylinder for each leaf, the two inner pursue a vertical 

 course into the petiole, while the remaining traces pass obliquely upwards into 

 the cortex and the leaf base, where they anastomose and form two girdles. 

 Girdling is characteristic of most cycads, but in the adult stem of Macrozamia 

 and in the seedling of Boivenia the trace bundles take a direct course — a more 

 primitive condition. The structure of the cycadean foliar bundle and its sig- 

 nificance in phylogenetic connections also attracted attention. Two opinions 

 opposed one another. One (Scott, 1923) was that this bundle is strictly mesarch 

 and agrees closely with that of Lyginopteris. Other anatomists (Chauvead, 

 1912) instead designated the cycadean foliar bundle as diploxylic on the ground 

 that the centrifugal and centripetal xylem were of independent origin and re- 

 mained distinct during most of their course along the petiole. 



The formation of a large number of spermatozoids in Microcycas aroused 

 great interest from the taxonomic viewpoint. Reynolds (1924) stated that this 

 genus has both primitive and advanced characters; that most of the former ap- 

 pear in the gametophytic generation, while most of the latter are in the sporo- 

 phytic generation; and that, since the advanced characters are more numerous 

 than the primitive, Microcycas should be regarded as one of the more advanced 

 cycads. 



Worsdell (1906) regarded the cycadean stele as derived from that of the 

 polystelic Medullosaceae. Scott (1923 and earlier), on the other hand, would 

 have the cycadean type of vascular anatomy derived from that characterizing 

 the Lyginopteridaceae. Matte (1904) believed that the cycads originated from 

 the latter through the Medullosaceae. In her study of Sutcliffia, de Fraine 

 (1912) agreed with Worsdell that their probable origin was along the medul- 

 losean line, but from monostelic (and protostelic) rather than from polystelic 

 forms. From the available evidence it seemed to Bancroft (1914) safe to pre- 

 sume that both pteridosperm families had arisen from a basal stock which also 

 had originated the cycadean line. The cycads were generally regarded as being 

 related to the mesozoic bennettites. Wieland (1906, 1916) retained both groups 

 in the Cycadales, notwithstanding great differences in the morphology of the 

 reproductive organs, but this view was later abandoned. 



