Ranales, Rhoeadales , and Rosales. 
73 1 
cannot regard the cotyledonary strands as the fixed quantity which, when 
present, will produce a certain type of root. That is to say, there is no 
inevitable connexion between the presence or absence of lateral strands 
and the possession of a tetrarch or diarch root respectively. The figures 
and diagrams in this paper demonstrate this point. When absence of free 
laterals is correlated with tetrarchy we have the so-called ‘ Anemarrhena 
type ; when laterals are present, although the root is diarch, they may 
perhaps be regarded as an indication of tetrarch affinity, since this feature 
is most commonly found in the allies of tetrarch species. 
The factors which control the freedom or aggregation of the lateral 
strands of the cotyledon petiole are probably capable of elucidation, 
and there is a fairly obvious correlation between this anatomical feature 
and the morphology of the cotyledons, particularly with regard to their 
insertion. The factors which control the production of a tetrarch or of 
a diarch root are more difficult of analysis. It is probably ftrue that there 
is a broad correlation between habit of seedling and root anatomy, but the 
Cruciform Type- Tetrarch variety 
a 
b 
C 
4 / ' 
\ / / \ 
\ . r\ \ / \ ; 1 
r 
( 
1 '/ J 1 « 1 I (r 
/ 0*0 \ 
( < V J 
\j ( y 
/ \ ) 
y \ / v / 
V O VO j 
Text-fig. 43. 
connexion is 
by no means close or universal. 
Thus comparison of the 
habit of the seedlings of the tetrarch and diarch species of Berberis and 
of the Magnolias reveals no obvious difference and it is interesting to note 
that among the Cactaceae the normal and comparatively slender seedlings 
are tetrarch, while the swollen much modified forms are diarch. 
Lee remarks upon this lack of direct correlation in the Compositae, and 
numerous other instances could be furnished. 
The root symmetry is undoubtedly affected by local as well as by 
more fundamental causes. Thus many authors have recorded cases of 
reduction from the tetrarch to the diarch condition in the apical region 
of the individual, and an increase in poles may also occur, but not so 
commonly. (See Hill and de Fraine on Classification of Seed-leaves. 1 ) 
The anatomy of the hypocotyl, however, is frequently practically 
1 Ann. Bot., vol. xxviii, 1914. 
