444 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



Juglandaceae, Myricaceae, Balanopsidaceae, Leitneriaceae, Fagaceae, Betula- 

 ceae, Corylaceae, and Salicaeeae, the Urticales, Rhoiptelea, and Proteales pos- 

 sibly being connected with them. A similar disposition is supported by Jan- 

 chen (1950). 



Thus, the questions whether "the Amentiferae" represent a natural group 

 or an artificial one containing unrelated families and whether they are primi- 

 tively simple, exhibiting direct connection with the gymnosperms, or their pres- 

 ent features are derived from ancestral forms with more complex flowers may 

 be tested in a relatively few families, as follows : 



1. CASUARiNACEAE. Casuarinaceae (Verticillatae) were placed at the very beginning 

 of the angiosperms by Bngler and by Wettstein; Rendle placed them as his sixth order. 

 Schnarf and Hjelmqvist have regarded the family as truly primitive on the basis of 

 embryological features presumably connecting them to gymnosperms. Lam places them 

 together with the Gnetales in the "Protoangiospermae," but "without any phylogenetical 

 meaning." Gaussen thinks they may stem directly from the Articulatae! Evidence 

 derived from stem anatomy (Bailey and Sinnott, 1914; Tippo, 1938; Moseley), floral 

 features (Moseley; Corner; Eames, 1951), and pollen (Moseley) seems to support the 

 view that "the Casuarinaceae are a specialized family of the Angiosperms and are not 

 a primitive group" (Moseley, 1948, p. 276), perhaps derived from Hamamelidaceae, as 

 suggested by Tippo, Copeland, and Moseley. 



2. FAGALEs. George (1931) stressed the similarity in vegetative features between 

 Gnetaceae and Fagaceae. The presence of aggregate rays, according to Hoar (1916), 

 suggested that Betulaceae and Casuarinaceae should both be placed low in the phylo- 

 genetic scale. However, evidence from stem structure (Bailey and Sinnott, 1914; Hall, 

 1952), the inflorescence (Rickett; Langdon, 1947), and the flower (Berridge, 1914; Abbe, 

 1935. 1938; Wilson and Just; Porsch, 1950) provides more than a hint that this is a 

 highly reduced group, with possible aflfinities to epigynous Rosales. 



3. JUGLANDALES. Hagerup (1938) considered the structure of the gynoecium as indi- 

 cating that Juglandaceae, together with Piperales and Caryophyllales, belong in the 

 same evolutionary line with Gnetales and Coniferales, and Hemenway (1911) regarded 

 the phloem of the family as relatively primitive. However, evidence from wood anatomy 

 (Heimsch, 1938; Heimsch and Wetmore), the inflorescence (Manning, 1938, 1940, 1948; 

 Rickett), and floral anatomy (Manning, AVilson and Just) appears to bear out the view 

 that the primitive Juglandales were plants with a paniculate inflorescence, bisexual 

 flowers with a prominent perianth, numerous stamens, several carpels, and possibly a 

 capsular fruit. The modern members of the alliance exhibit a high degree of reduction. 

 Withner (1941) considered Rhoiptelea to represent a relatively primitive member of this 

 order. Convincing data have been gathered together by Withner, Heimsch (1942), 

 Hjelmqvist, and Stern (1952) to show that Julianiaceae do not belong either in or near 

 Juglandales, but possess affinities rather with Anacardiaceae. 



4. SALK'ALEs. Spomc, Consistent with his contention that angiosperms had primitively 

 unisexual flowers, regards the group as "quite primitive," and von Tuzson (1936) thought 

 Salix one of the most archaic of dicots. However, evidence from the structure of the 

 wood (Holden, 1912; Eames, 1951), the nature of the inflorescence (Fisher, 1928), floral 

 anatomy (Fisher, Eames, Wilson and Just; Nagaraj, 1952), and embryology and cytology 

 (Wilkinson, 1944; Nagaraj) appears to establish a strong inference that poplars and 

 willows "though doubtless belonging to one of the more primitive lines of angiosperms, 

 are far from primitive" (Eames, 1951, pp. 30-31). Even if the view is accepted that the 

 order is a derived one, it is not clear what relationships it exhibits to other angiosperms: 

 Bessey related it to Caryophyllales, Brown and Gunderson to Parietales or possibly 

 Rhoedales. 



5. URTICALES. Evidence for the advanced rather than primitive status of Urticales is 

 based on the anatomy of the stem (Chalk, 1937; Tippo, 1938, 1940) and the flower 

 (Bechtel, 1921; Eames, 1926; Eckardt). A relationship is suggested by Tippo, Copeland, 

 and Moseley to Hamamelidaceae. 



