190 
‘Following the diagnosis of Peglera capensis and the remarks on 
the«systematic position of this species in the Kew ‘Bulletin (1907, 
pp. 362-3), a°short description of the «anatomical ‘structure of ‘the 
plant was given,'and “a comparison was made with’ the anatomical 
characters of ‘the Legnotideae (Rhizophoraceae) and the Simaru- 
baceae. ‘The conclusion ‘arrived at was that the genus ‘Peglera 
“would ‘find a more natural place in‘ the Simarubaceac than in’ the 
‘Leynotideae.”’ ae 
A comparison’ has now' been’ made! between ‘Peglera capensis and 
Nectaropetalum Kaessneri, and the agreement of ‘the “anatomy m 
these ‘two plants :proves ‘to be sufficiently marked to establish a 
‘fairly close affinity between’ them. ; : 
There-examination of Peglera, for the purpose of this comparison, 
has ‘led ‘to ‘the detection of two characters, which’ had been over- 
looked:at' the‘ time when’ the: previous anatomical description of ‘the 
genus was written. One of these is‘the presence of occasional 
‘scalariform ‘perforations in ‘the vessels, simple perforations being 
general in'the-wood, and those of ‘the scalariform rare. ‘The othe 
character is ‘the oceurrence of ‘cortical vascular bundles whieh, 
however, are restricted to a small portion 6f the ‘internode yust 
below the node. 
The agreement between P. capensis and N.‘Kaessneri extends sto | 
most ofthe lesser-anatomical details, while of the points of difference 
the chief are-as'follows : vascular:system of midrib slightly-simpler 
in Nectaropetalum ; papillae on the lower epidermis: of the leaf im 
Nectaropetalum but not in ‘Peglera ; ‘cortical bundles -extending 
throughout the internode:in Nectaropetalum, but oceurring only near 
the top*of:the internode jin “Peglera. The first two differences*are 
quite: unimportant, and ‘the third, though considerable (see below), 
is one that would not be very surprising if found within the limits 
of acgenus, 
‘In Neetaropetalum two vascular’ bundles separate from the*stele'1n_ 
the ‘upper part ‘of ‘the internode “and pass ‘upwards, as cortical 
rae Sarna the node ‘and the next internode. ‘They reaeh the 
“secon 
e, 
and four'in the upper part. Thus the behaviour of the cortical 
r both may 
separate from the stele just before or just after the two cortical 
bundles from the mode below pass out of ‘the stem.§ Hence, 4§ 
* The material examinéd was insufficient for d ini lation of the 
. etermining the relation 
cortical bundles to the Stipules, &c., with certainty 
Van Tiegh aed fc das: 
ix (1908), ig” _ Struct. et affin. des Erythroxylacées, Bull. Mus. dhist 
¢ Specimen from Namasi, leg. Cameron, No. 3. 
me this case it was determined with certainty that each of these two bundles 
tks . the node, giving a small branch for the stipule, and a larger ie ands 
fuses with one end of the median leaf-trace, as described by Van Tieghem- 
