298 MR. SPENCER LE M. MOORE—PHANEROGAMIC BOTANY 
layer. The medullary rays vary considerably in width; like the cells of the com- 
paratively small pith-region, they are filled with spherical or elliptical simple or more 
usually compound starch-grains, varying greatly in size, the largest being usually found 
in the pith. 
The leaves are inserted on a strictly distichous plan, but material for the study of the 
arrangements of the leaf-traces is not available. The one point worthy of mention à 
propos of the leaf-structure is that the stereome protecting the vascular bundles runs 
right up to the epiderm, which latter tissue, where it overlies the bundles, is composed 
of smaller cells than usual, and of cells often with somewhat thicker walls than else- 
where; in other words, the epiderm itself can to a certain extent assume the appearance 
of stereome, and this arrangement must obviously give great strength to the leaf. 
Beneath the midrib, however (that is, on the lower side of the leaf), parenchyme inter- 
venes between the stereome and the epiderm (fig. 3). Fig. 2 shows the upper surface of 
the leaf seen from above; over the palisade tissue the epiderm cells are larger and have 
the wavy borders so frequently possessed by epiderm; but the cells lying over the vascular 
bundles are shorter and more angular, and usually, though not quite always, each of 
them contains a single rhomboidal crystal. The leaves are bifacial in the most marked 
degree, there being no stomates on the upper surface; each stomate is accompanied by 
two subsidiary cells, often very unequal in size (fig. 4). Near the vascular bundles the 
mesophyll seems to be continuous, without any intercellular spaces; further away, 
however, it becomes “spongy.” Many Anonacee have glandular leaves, and 
Ephedranthus is no exception to this, as sections show several glands immersed in the 
lower part of the mesophyll. These glands are similar in appearance to those of 
Stormia, shown in fig. 7 of Plate XX XVIII. 
Like the leaves, the three series of bracts are strictly distichous. In the case of one 
flower, however, which I examined, the sixth bract did not stand exactly over the fourth, 
but had been deflected to a trifling extent towards one side, probably as a result of 
abnormal pressure in the bud. In the diagram (Plate XXXVITI. fig. 5) it will be seen 
that the position of the axis is not indicated; this I have not been able to ascertain, 
owing to want of material. 
The small pollen-grains are ellipsoidal in shape, and, except for a few fine striations, 
are quite smooth. I could make out two pores, but never more. 
GUATTERIA SYLVICOLA (sp. nov.) ; caule subtereti, leviter undulato, nigrescenti-purpur- 
.  ascente rufo-tomentoso, dein pubescente, demum glabro ; foliis maxime inzequalibus, 
subsessilibus, elongatis, anguste oblongo-obovatis, cuspidatis, obtusis, basi cuneatim 
angustatis, coriaceis, supra fere glabris, subtus, preesertim secusnervos, rufo-tomentellis ; 
pedunculis solitariis, foliis longioribus, prope basin articulatis, indumento rufo-villoso- 
tomentoso instructis; bracteis minutis ; sepalis late ovatis, basi paullo cordatis, 
obtusiusculis; petalis obovato-oblongis, obtusis vel retusis vel brevissime bilobis, 
sepala duplo excedentibus ; staminibus late linearibus obtusissimis; baccis ignotis. 
Hab. In sylvis ad Serra da Chapada, alt. circiter 2000 ped. supra mare. Fl. Aug. 
(N. 142.) 
