762 TRANSACTIONS OF SECTION K. 



a break in the history of development of the stem, -which masks, if it does not 

 destroy, the succession of phases of vascular evolution met with in Ferns. 



Nevertheless the hypocotyl itself, with its direct continuation downwards — the 

 primary root — and its appendicular organs — the cotyledons — may be considered 

 apart from questions as to its origin, so that it remains to determine whether its 

 characters, such as the anatomical features of its vascular system, have phylogenetic 

 value ; and if so, what are the actual plans of structure met with, and the phases of 

 evolution through which they pass. 



Miss Sargant's worli ^ has established the phylogenetic value of the vascular 

 symmetry of the hypocotyl in certain Monocotyledons, particularly Liliaceae, and 

 our results for Gymnosperms and Dicotyledons confirm and extend her conclusion. 



The difterent types of ' transition ' between the collateral structure of the vas- 

 cular tissue of the cotyledons and the radial structure of the primary root-cylinder 

 we find to be all, with a very few exceptions, modifications of one great plan. 

 The primary root is, in the vast majority of cases, either diarch or tetrarch, 

 and the plane passing through the centres of the two cotyledons (cotyledonary 

 plane) also passes through the protoxylem poles of the diarch root, or through two 

 out of the four poles of the tetrarch root. This was established by Dangeard - in 

 1889. 



The characteristic dicotyledonous type has. near the base of each cotyledon- 

 petiole, a striking tj'pe of bundle — the ' double bundle ' — very variable in details, 

 but characteristically consisting of a V-shaped strand with two pliloem groups on 

 the arms of the y, and the xylem occupying the apex. Frequently this bundle is 

 exarch, i.e., the protoxylem is turned outwards between the phloems. At or 

 shortly below the cotyledonary node the xylems of the two cotyledonary double 

 bundles become exarch, if they are not so already, and fuse to form the diarch plate ; 

 while of the four phloems two join on each side of this plate to form the two 

 phloem strands of the root-cylinder. This type is found, not only in the Ranun- 

 culaceiB and several allied families and throughout the Rhseadales, but also 

 in many other families of Archichlamydese, and is apparently typical of the 

 Sympetalfe. 



In addition to the ' double bundle ' two laterals exist in many cotyledon - 

 petioles. Often they join on to and lose themselves in the double bundle towards 

 the base of the petiole, but in other cases they may persist into the node and each 

 take up a position, close to the corresponding one from the other cotyledon, in 

 the intercotyledonary plane. Their xylems may either eventually lose themselves 

 in the sides of the diarch plate, the protoxylems dying out {Liriodendran, Ci/cas 

 Siamensis), or they may form the intercotyledonary poles of a tetrarch root 

 {Gasuarina, AltJicea, Ceratonia, &c.). 



In other cases, again, in which a tetrarch root is formed, we have exactly the 

 same state of things as that just described, except that the double bundle is repre- 

 sented in the cotyledon-petiole by two distinct and separate collateral bundles 

 (making, with the laterals, four in all), which eventually rotate, with their xylems 

 inwards, so that their protoxylems come together to form the cotyledonary pole 

 of the root-cj'linder (Hicinus, Ciicurbita, Acer, &c.). 



It is plain that the cases described are naturally regarded as stages in a series. 



The evidence points, we think, to the last case described being phylogenetically 

 the oldest type, since it is typical (so far as we have been able to extend our researches) 

 of Oycads and of Araucariese. It has been retained, we think, in certain dicotyledons, 

 though the transition from the shoot-type to the root-type of vascular arrangement 

 has often been shifted right down to the base of the hypocotyl or even into the 

 primary root. Most of the Coniferse and the vast majority of Dicotyledons investi- 

 gated by us have, however, undergone, according to our view, a reduction, first by 

 the fusion of the two central cotyledonary bundles to form the * double bundle,' 

 the tetrarch type of root being here retained, and eventually by the disappearance 



' ' A Theory of the Origin of Monocotyledons based upon the Structure of their 

 Seedlings,' Ann. of Bot., 1903. 



^ Dangeard, ' Kecherches sur le mode d'union de la tige et de la racine chez les 

 Dicotyledones,' Le Botanute, 1889. 



