ANATOMY OF THE ROOTS OF pAlys- 447 
| bifurcation would be explicable as resulting from two large Jateral roots arising close to 
| the end of the principal one of the nature of those in Areca Oglechy (P). 51, fig. 116) ; 
but such an explanation is here quite without support, 2nd dichotomy seems to afford 
the only satisfactory explanation of this peculiarity, 
The structure of the series of sections represented by pl. 59, figs. 55 & 76 may now be 
explained in the light of the facts established by the examination of Kentia, Latania 
borbonica, and Dictyospermum album. Each section represents thé most complex 
condition met with in the extracauline portion of the root uy der consideration. 
The section of Areca sp. (Pl. 50. fig. 76) illustrates the most complicated root- 
structure met with in the Palms. The original cylinders of (he pase of the root become 
united to form not a single ring, but three complete eylingers each enclosing a “ pith” 
composed of ground-parenchyma, and a large incomplete cylinder Whose internal portion 
is made up of three of the original strands each preserving its radial structure. The 
complete cylinders show the internal lobing of the fibrous ring jn the form of wedges, 
each representing an originally separate strand from whieh the internal phloem has 
disappeared and also internal protoxylem, leaving onlY a large vessel, sometimes 
accompanied by one or more smaller elements, to represent phe petaxylem of originally 
centrally orientated xylem-bundles. 
In the incomplete cylinder in one case an interpally dipecteg ProtoXylem-group 
persists, and in several cases two or three of the metüXy]em.eJements of these 
bundles remain. 
The radial symmetry of the fibrous groups forming the edges of the incomplete 
cylinder exhibit a nearly complete radial structure, which is also exhibited by the three 
strands forming the internal portion of this cylinder. 
In the parenchymatous tissue lying within this cylinder the medullary groups show a 
mass of xylem with two or three radiating protosylemSrotbs aad the same number 
of phloem-groups, the original radial strands not having Jost their Primitive structure, as 
is usually the case with medullary strands. M 
In Jriartia (Pl. 50. fig. 55) the peripheral strands have vie Into à series of ares, 
the internal lobing of which indicates the boundary of the oríSinally free strands. The 
internal protoxylem-groups have aborted and generally also the phioem-groups, but in | 
the segments forming the free edges of the arcs an interna] phloem-proup often persists, 
The medullary strands preserve their original radial sy™Mel"y, usually possessing two 
or three xylem-groups alternating with as many phl?mStouPs, The endodermis 
bounds each are externally and is continued inwards Youd the edges for a short 
distance, but is soon lost as a definitely marked ]ayey In the common ground- 
arenchyma. : 
Areca sp. (PL 50. fig. 56) is essentially similar to Zria"tia, but the medullary strands * 
have lost their phloem-groups. The wedge-shaped interna] Jobes of the fibrous tissue of 
the ares are well-preserved and an internal phloem-gro"P i$ CopMonly present in the 
lateral lobes of any are. ? i 
In Kentia sp. (Pl. 50. fig. 57) the fusion of the ates ÍS soMewhat Further advanced, 
The wedges are still well-marked, but medullary strands afp apsent, Al] (he original strands 
