416 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



with type of ray was attempted by Wetmore (1926), who regarded transverse 

 lenticels as primitive in comparison with longitudinal ones. Later, Kribs (1937) 

 used the same statistical approach to establish a sequence of the types of dis- 

 tribution of wood parenchyma, ranging from apotracheal, or diffuse and meta- 

 tracheal, to the more specialized paratracheal or vasicentric types. Gilbert 

 (1940) asserted that the advanced condition of ring-porosity occurs only in the 

 north temperate zone and without any correlation with the ray types defined 

 by Kribs. It seems clear that the form and distribution in the wood of rays and 

 parenchyma should not be used alone in any endeavor to construe stem anatomy 

 phylogenetically. 



Chalk found it possible to classify all dicotyledonous woods into three groups 

 of different degrees of specialization on the basis of the occurrence of vessels 

 with scalariform perforations, fiber-tracheids, and storied structure of the 

 wood. The same methods were extended to the woods of fossil dicots, and his 

 observations tended to show "that the characters regarded as unspecialized by 

 the wood anatomist were relatively more common in the past" (1937, p. 423). 

 The close similarity of such woods as the fossil Homoxylon to both cycadophytes 

 and vesselless Ranalians has been used as indicating either direct derivation of 

 the angiosperms from, or their close affinity with these gymnosperms (Sahni, 

 1932, 1935; Wieland, 1933, 1934; Gupta, 1934; Hsii and Bose, 1952). 



The basic sequences established by Frost and Kribs have been correlated 

 with a number of other features of wood anatomy, so that a continually grow- 

 ing list of phylogenetically significant wood characters has been made available 

 for purposes of taxonomic comparison. These have been summarized by Vestal 

 (1937, 1940), Tippo (1938, 1946), Moseley (1948), Cox (1948a, 1948b), and 

 Hall (1952). The intensive application of anatomical criteria to different groups 

 of woody dicotyledons either for clarification of external affinities, or for better 

 classification of constitutent groups may be exemplified by the following : Grui- 

 nales and Terebinthales (Kribs, 1930; Webber, 1936, 1941; Heimsch, 1940, 1942, 

 Stern, 1952), Guttiferales and Parietales (Tupper, 1934; Vestal, 1937; Taylor, 

 1939); Juglandales (Heimsch, 1938; Heimsch and Wetmore, 1939; Withner, 

 1941), Malvales (Chattaway, 1932, 1937; Webber, 1934), Eanales (Garratt, 

 1933a, 1933b, 1934; McLaughlin, 1933; Bailey and Smith, 1942; Bailey, Nast 

 and Smith, 1943; Bailey, 1944a; Bailey and Nast, 1945a, 1945b, 1948; Lemesle, 

 1946a, 1946b, 1946c, 1948, 1953; Nast and Bailey, 1946; Bailey and Swamy, 

 1948, 1949; Swamy and Bailey, 1949, 1950; Swamy, 1949; Money, Bailey and 

 Swamy, 1950), Urticales (Tippo, 1938, 1940), Betulaceae (Hall, 1952), Casuari- 

 naceae (Moseley), Ericaceae (Cox), Icacinaceae (Bailey and Howard, 1941), 

 and Lecythidaceae (Diehl, 1935). Lists of supplementary features suitable for 

 comparative study and identification, but without phylogenetic seriation, have 

 been supplied by Eecord (1936), Record and Chattaway (1939), Metcalfe 

 (1946), and Metcalfe and Chalk. Metcalfe and Chalk's monumental compen- 

 dium of anatomical data, like that of Solereder before it, provides information 

 on most dicotyledonous families and a very rich bibliography. It would perhaps 

 not be out of place to remark, however, that such data are as yet extremely 

 fragmentary for the majority of plant families. 



There is some evidence to suggest that the development of vessels in primary 

 xylem has paralleled that of those in the secondary xylem (Bailey, 1944b). 



