FLORIN: SYSTEMATICS OF THE GYMNOSPERMS 343 



while the other regarded the microsporophyll of the Pinaceae, with two spo- 

 rangia on the dorsal surface, as the basic tyi)e. Dupler (1919) held that the 

 peltate, perisporangiate microsporophyll had ])robably been carried forward to 

 modern gymnosperms by the cordaitalean line. The araucarian microsporophylls 

 suggest a basic peltate structure, and true i)eltate forms occur in Taxus and 

 Torreya (Coulter and Land, 1905), though in the latter genus the adult sporo- 

 phyll becomes hyposporangiate. Doyle (1926) believed that the primitive pel- 

 tate sporophylls were never by nature foliar, but sporangioi)horic. Pilger (1926, 

 1929), on the other hand, comprehended the microsporophyll as a metamor- 

 phosed leaf, and regarded as primitive such conifer sporophylls as only differed 

 slightly from ordinary leaves. The microsporophyll of Taxus, however, was 

 assumed to be characterized by a morphogenesis of its own. 



Stimulated by the European activities in fossil botany, the wood anatomy 

 of the gymnosperms developed into an important subject. Diverse opinions on 

 the phylogenetic meaning of many observed structures developed, however, and 

 conifer anatomy became a ground of debates of the first order. The type of 

 tracheary pitting in secondary xylem was held to be of phylogenetic interest. 

 Gothan (1905) considered that the most primitive type of bordered pitting was 

 hexagonal, alternate, and crowding the whole tracheid wall (araucarian pit- 

 ting). Elimination resulted first in small isolated groups of bordered pits, then 

 in the uniseriate flattened condition, and finally in the scattered arrangement, 

 where the pits occur singly or in opposite arrangement on the radial wall (mod- 

 ern pitting). The transitions between the two types found in fossil woods of 

 mesozoic age were believed by Cxothan to represent relatively primitive extinct 

 Pinaceae. The Araucariaceae were the most ancient conifers, and derived from 

 the cordaites. Jeffrey (1912), on the other hand, looked upon the araucarian 

 wood as more recently acquired, and based this conclusion on the structure of 

 the first annual ring of mesozoic Araucarioxylon stems, of the seedling of living 

 araucarians, and especially of the wood of their cone axes. He regarded the 

 Pinaceae as the ancient and primitive conifer group, most closely associated 

 with the cordaites, and the Araucariaceae as derived from it. Thomson (1913) 

 emphasized the resemblances of the Araucariaceae to the Cordaitales in the pit- 

 ting of the secondary tracheids in the root and in the axis of the female cone. 

 Jeffrey considered that the presence of rims of Sanio in the secondary tracheids 

 adjacent to the primary wood of the cone axis in Araucaria and AgatJiis sup- 

 ported his view of the pinaceous ancestry of the araucarians. But Thomson in- 

 terpreted the rimlike thickenings and alternate pitting sometimes found in the 

 cone axis and first annual ring of stems and roots in the pines as indicating the 

 opposite. In Bailey's (1919; cf. Pool, 1929) view, it is in both cases a question 

 of a transitional type of tracheary pitting and not true crassulae. Sifton (1920) 

 agreed with Bailey's statement that neither of the two argumentations is ten- 

 able. The taxonomic importance of the crassulae was very differently evaluated. 

 Jeffrey (1912), and his associates, accepted them as an infallible criterion for 

 diagnosing coniferous woods, while Gothan (in Potonie and Gothan, 1921) con- 

 sidered that Jeffrey's school exaggerated their importance. Bailey (1925) as- 

 serted that the conifers are, in contradistinction to the cordaites, characterized 

 by the circular bordered pits having worked back into the earlier formed por- 

 tions of the primary xylem, and by the elimination from the transitional zone of 



