45i 
Seedling Structure of Gymnosperms. III. 
The behaviour of the cotyledonary traces of this species is obviously 
different to that of the same structures of E. Altensteinii , and we have seen 
nothing exactly like it in other genera. It is interesting to find that 
Karsten 1 has described each cotyledon of Zamia muricata as receiving one 
vascular trace from the axis which branches in the higher regions of the 
seed-leaves. This apparently does not obtain in all species of Zamia , for 
Miss Dorety, in her paper on Ceratozamia , 2 states that some species of 
Zamia are of the same type, as regards their structure, as Dioon edule. To 
return to E. Barteri , the later vascular rearrangements leading to the 
formation of the root-structures apparently are similar to those of Macro - 
zamia spiralis. 
The following descriptions deal with plants which we have not had the 
opportunity of examining. In many cases the authors cited have followed 
in their descriptions of the transition-phenomena a method different to that 
pursued by us ; in such instances we have described the facts in our own 
way for the sake of uniformity. 
Cycas. 
Cycas siamejtsis, Miq. According to Matte 3 the transition-phenomena 
are similar to those described by him for Dioon edule with this difference, 
that the lateral cotyledonary bundles do not form any intercotyledonary 
poles, so that the resulting root-structure is diarch. As regards Cycas in 
general the following paragraph is quoted from Miss Thomas 4 as it 
embodies a summary of some of the facts observed by her and Tansley . 5 
‘ It appears that Cycad cotyledons are almost without exception 
destitute of a midrib, and have four or six bundles at their base ; that the 
transition features take place so near the cotyledonary node that the hypo- 
cotyl is root-like in structure ; and that the primary root is in most cases 
tetrarch, with a strong tendency to reduce near its apex to diarch. The 
genus Cycas would appear to be most aberrant, for while C. revoluta forms 
a tetrarch root, C. siamensis . . . forms a link between C. revoluta and 
C. Rumphii , which is diarch through the absence of intercotyledonary poles. 
Further, C. revoluta may have its two central bundles replaced by a single 
one, which, however, plays the part of two.’ 
There is apparently some variation, for Bower G states that in Cycas 
Seemanni there may be in the cotyledon a median bundle, or two equal 
ones disposed symmetrically near the centre of the cross section, and be- 
tween these extremes intermediate modes of arrangement may occur. He 
1 Karsten: Organographische Betrachtungen der Zamia muricata (Abh. d. Berlin. Akad., 193. 
1856). 2 Bot. Gaz., xlvi, 1908. 
3 loc. cit. 4 loc. cit. 5 British Association, Section K, York, 1906. 
6 Bower : On the Comparative Morphology of the Leaf in the Vascular Cryptogams and 
Gymnosperms (Phil. Trans. Roy. Soc., Lond., 175. 1884). 
