CONSTANCE: SYSTEMATICS OF THE ANGIOSPERM$ 439 



of seed evolution" (1946, p. 529). There appears to be some correlation between 

 a parietal (peripheral) embryo and starchy endosperm, and a central (axile) 

 embryo with relatively nonstarchy endosperm, if any. Subordinate to the two 

 divisions are twelve "types." Martin suggests a few trends which he regards as 

 evolutionary, beginning with primitive medium-sized or large seeds containing 

 relatively small embryos and manifesting early dormancy. Two trends in size 

 are evident, the one leading to quantity production of minute and delicate seeds 

 (microspermy), the other to the formation of a few large, well-stocked, relatively 

 large-embryoed ones (megaspermy). A study of the "family tree" is very in- 

 structive, although the author warns that it is not meant to represent a new 

 classification of families for general purposes, or to supersede data from other 

 lines of evidence. There appears, for instance, to be a close correlation in seed 

 type between Nymphaeaceae and Saururaceae with certain monocotyledons; 

 Cyperaceae and Gramineae manifest important differences; seeds with a curved 

 peripheral embryo — Caryophyllales, Cactaceae, Frankeniaceae — are regarded as 

 an evolutionary blind alley. Although such an angiospermous "seed phylogeny" 

 is obviously premature, it seems to indicate that there do exist characters and 

 trends which might be employed not only for taxonomic but also for phylo- 

 genetic objectives. 



COMPAKATIVE BIOCHEMISTRY 



The idea that comparative biochemistry, like comparative morphology, anat- 

 omy, and embryology, can furnish significant evidence for relationship and clas- 

 sification is attractive and probably theoretically sound. Indirectly, such chemi- 

 cal features as the presence of latex, resins, volatile oils, and the possession of 

 different kinds of endosperm have long been employed in the recognition of 

 members of certain families and genera, and even as suggesting affinity between 

 families. 



The distribution of anthocyanins in flowers, it has been suggested, possibly 

 has "some phylogenetic significance, although there are obvious limits to the 

 conclusions which may be drawn" (Lawrence et al., 1939, p. 173). The posses- 

 sion of nitrogenous anthocyanins affords an additional agreement between Cac- 

 taceae and families of Caryophyllales (Gibbs, 1945). The occurrence of fats and 

 fatty acids in plants "runs on the whole remarkably parallel with the groups into 

 which morphologists have placed them" (Hilditch, 1940, p. 14). However, "the 

 biogenesis in plants of fats from carbohydrates remains indeed an uncharted 

 and mysterious field" (Hilditch, 1952, p. 182). Potentially, the comparative bio- 

 chemistry of essential oils, as applied by Baker and Smith (1920) to Eucalyptus, 

 holds considerable promise (McNair, 1942). Plant alkaloids which are specific in 

 their occurrence may also be of classificatory value since, according to Weevers, 

 "a close relation exists between the nature of the chemical products and the 

 taxonomical position of the species, genus or family which produce them" (1943, 

 p. 421; McNair, 1935a; Weevers, 1933). 



In his exhaustive monograph of starches and their reaction-curves Reichert, 

 although acknowledging "the very limited range and preliminary nature" of his 

 research, emphasized the worth "of the molecular characters of products which 

 are passive, nonstructural constituents of the plant" (1913, p. 340). He also gave 

 examples to suggest the desirability, on this basis, of subdividing the families 



