FLORIN: SYSTEMATICS OF THE GYMNOSPERMS 351 



seiilcd oi' gymnospenns. Various asj)e(,:ls ui' (le\clupiuenlal mori)liulo8y were dis- 

 cussed by Sifton (1944). 



The results of the researches into the origin and development of primary 

 vascular tissues in seed plants were reviewed by Esau (1943). Bailey continued 

 his investigations of the structure of the cambium and its derivative tissues, and 

 clarified to a considerable extent the problem of the cell wall structure of liigher 

 plants (Bailey, 1940). He suggested that the old and still open question of the 

 mode of formation of the secondary wall — wiiether by intussusception or by 

 apposition — wall ultimately be solved essentially in favor of the latter alterna- 

 tive. Other workers interested themselves in tlie origin, development, and distri- 

 bution of xylem rays in gymnosperms. According to Bannan (1934) the evidence 

 of the phylogeny and of the ontogenetic sequence indicates that the primitive 

 ray is parenchymatous, and that ray tracheids have arisen at the expense of 

 parenchyma. Esau (1939, 1950) also summed up the work done on phloem tis- 

 sue. This kind of research had come almost to a standstill about the end of the 

 nineteenth century, and was not revived until more tlian thirty years later. Pith 

 structure, particularly in conifers, had been even more neglected until Doyle 

 and Doyle (1948) began a series of works in this field. The origin of trans- 

 fusion tissue in the leaves of cycads. Ginkgo, and conifers was again brought up 

 for discussion. Van Abbema (1934) held that Mohl's (1871) theory that the 

 central transfusion tissue is nothing but modified parenchyma and thus of the 

 same nature as the accessory transfusion tissue, was likely to be correct, while 

 Huber (1948) supported Worsdell's theory of 1897. The marked advances of 

 late years in our knowledge of gymnosperm anatomy have been summarized by 

 Eames and MacDaniels (1947) and Foster (1949). The structure of the epi- 

 dermis was, however, only cursorily treated. Florin (1931, 1933) studied the 

 epidermal characters of the recent gymnosperms from the taxonomic point of 

 view, and found that external leaf morphology and epidermal structure — with 

 special reference to the structure of the stomatal apparatus — constitute a feature 

 complex which is generally well suited to serve as a means of characterizing 

 natural species groups of generic rank. 



In the pteridosperms and the cycads the guard cells of the stomatal appara- 

 tus are directly originated by the primary mother cell. The surrounding (peri- 

 gene) epidermal cells may function directly as subsidiary cells, or each may 

 divide into one subsidiary and one or more radially arranged encircling cells. 

 This primitive haplocheilic or simple-lipped type also characterizes the cor- 

 daites, ginkgophytes, conifers, taxads, and ephedras. In the bennettites, on the 

 other hand, the primary mother cell of the guard cells usually divides into three 

 cells, of which the median cell gives rise to the guard cells, and the tw^o (meso- 

 gene) lateral cells function as subsidiary cells. One or both of the lateral cells 

 may also divide into one subsidiary and one encircling cell. This is the syndeto- 

 cheilic or compound-lipped type of stomata of the gymnosperms, which — apart 

 from the bennettites — occurs only in the living genera Welivitschia (Florin, 

 1934) and Gnetum. The haplocheilic type of stomata in gymnosperms is primi- 

 tive and the syndetocheilic type advanced. This accords with the fact that all 

 paleozoic gymnosperms so far examined have haplocheilic stomata, while the 

 syndetocheilic type does not appear until mesozoic time. The mode of develop- 

 ment of the stomatal apparatus thus constitutes a character sometimes separat- 



