9 2 
M. C. Stapes on 
nation, they present some features of interest; whilst the state 
of preservation is unusually good, especially in those tissues which 
generally leave much to be desired in this respect. 
The general type of the leaf, which is that figured by M. 
Renault under the name C. principalis , is shewn in phot. 2, plate 
IX. It is dorsiventral in structure, having the vascular bundles 
arranged in one plane, at an average distance apart of *45 mm. 
Each bundle is enclosed in a double sheath and accompanied by 
strands of sclerenchyma above and below. Large strands of the 
same sclerenchyma alternate with the vascular bundles on the lower 
face, whilst above there are three or four small and irregular 
groups just below the epidermis. 
The majority of the bundles are parallel and equidistant, and 
the branching, which appears to occur but seldom, is dichotomous 1 , 
the two strands moving very slowly apart. When commencing to 
divide, the phloem is separated into two groups by an intrusion 
of the inner sheath, the protoxylem remaining as one (see phot. 
4.); the protoxylems then split and move slowly apart, the inner 
sheath attaching itself to the centripetal xylem before either the outer 
sheath or sclerenchyma above shew signs of division. This 
behaviour is more striking than in the case of the recent Cycads, 
c.g., Zaniia or Ceratoza/nia, where the veins are similarly arranged, 
and in which the phloem also divides first into two free groups, 
but the above-described behaviour of the inner sheath is less 
pronounced. In Myeloxylon radiatum , the petiole of a Medullosa , 
where the individual bundle is of similar organization, the division 
appears to originate sometimes in the phloem and sometimes in the 
xylem. 
The parenchyma of the leaf is differentiated into palisade 
above and spongy parenchyma below. In sections in which the 
bundle is cut transversely this distinction does not appear very 
striking, the palisade cells are almost hexagonal and closely 
packed, while the spongy parenchyma appears similar but rather 
more irregular or lacunar. Between the two on each side of the 
bundle are a few clear cells with thin walls, elongated transversely, 
and although hardly specialised enough to be called accessory 
transfusion-tissue, they appear in some cases to merge into the 
cells with pitted walls at the sides of the bundles. 
1 Cf. Lignier, “ La Nervation des Cycadees est dicliotomique,” 
Assoc. Fran<;aise pour ravancemeiit des Sci., Congres de Caen, 1894, 
where lie demonstrates the fact that the branching of all the 
nerves in Cycads is truly dichotomous. 
