352 A CENTURY Of PROGRESS IN THE NATURAL SCIENCES 



ing taxonomic groups of high rank. It has become of special importance in con- 

 nection with the classification of fossil leaves of cycadlike type in deposits of 

 mesozoic age. Also Harris (1932-1937, 1942-1952) has investigated the stomatal 

 structures in cycads, living and extinct, as well as in bennettites, fossil ginkgo- 

 phytes, and conifers, etc. The stomata of the ginkgophytes were subjected to fur- 

 ther study from the taxonomic point of view by Florin (1936a), who, in addi- 

 tion, in a still later work (1938-1945, 1951) described the epidermal structures 

 of the leaves of the oldest known conifers of paleozoic age, and compared them 

 with those of the cordaites. Orr (1937) tested the value for diagnostic purposes 

 of these structures in living conifers in general, and Cookson and Duigan (1951) 

 in recent and fossal Araucariaceae. 



Research devoted to the gametophytes has continued in the modern period, 

 although with less intensity than that characterizing the first thirty years of 

 the present century. Its aim has been to fill remaining gaps in our knowledge 

 of their development and organization. Doak (1932) confirmed Juel's (1904) 

 observation that in Cupressus the pollen tube often develops a complex of sev- 

 eral male cells instead of the usual two, and considered this feature to be a 

 reversion. Florin (1936b) studied the structure of the male cordaitean gameto- 

 phyte at the shedding stage of the pollen grains. The central body of the grain 

 was not filled by w^alled cells, as Renault (1879) had believed, but appeared 

 possibly to have a peripheral layer of such cells. The interior of the body con- 

 tained a central row of free nuclei orientated along the vertical axis of the pol- 

 len grain. No pollen tubes have so far been found in paleozoic gymnosperms. 

 Summaries of the comparative cytology of the sexual apparatus and of the evo- 

 lution of the archegonium were published by Schnarf (1941, 1942). Regarding 

 the development of the male and female gametes, he emphasized their formation 

 in pairs. This "Zweier-Gesetz" is always modified in the female sex, however, 

 and sometimes also in the male, one of the two gametes having degenerated or 

 assumed a special function. General discussions of the evolutionary trends of 

 the gym.nosperm gametophytes were given by Fagerlind (1941) and Battaglia 

 (1951). 



A pronounced feature of the period under review is that to a large extent 

 the interest in "life-histories" of gjonnosperms shifted from the development 

 of the gametophytes and proembryo to their embryogeny as a whole. Johansen 

 (1950) has recently summarized our knowledge of gymnosperm embryology 

 with a view to facilitating correlation and evaluation of the results obtained. 

 Buchholz (1933) distinguished two kinds of cleavage polyembryony in conifers, 

 determinate and indeterminate. In determinate cleavage polyembryony one 

 embryo is more favorably situated than the others, and ordinarily becomes the 

 successful embryo, while in the indeterminate condition any one of several em- 

 bryos derived from the same zygote may survive. According to him, the probable 

 steps in the evolution of polyembryony were: (1) indeterminate cleavage poly- 

 embryony; (2) determinate cleavage polyembryony; (3) simple polyembryony 

 showing definite faces of determinate cleavage polyembryony; and (4) simple 

 polyembryony without any such traces. The phylogenetic theories developed in 

 conifer embryogeny by Buchholz have not been undisputed. In discussing the 

 Podocarpaceae, Doyle and Looby (1939) stated that the simple embryogeny of 

 Stacky carpus, Saxegothaea, and Phyllocladus appeared to be basal in the family, 



