SPADICES AND SPATHES. 3 
of D. fissus contained ovaries in process of development, which had every appear- 
ance of having been already fecundated. ‘The closure may however have been more 
apparent than real, and thus an insect may have been able to penetrate into the 
interior through the half-closed spathes. In the Daemonorops two very distinct forms 
of spadices may be distinguished, namely, that proper to the species belonging to the 
section Cymbospatha and that belonging to the section Piptospatha. 
In Cymbospatha the spadices, unopened or still completely enveloped by the 
spathes (d and $2), are at the moment preceding anthesis fusiform, more or 
less thick in the centre and prolonged into a kind of elongated beak, which 
at times exceeds the body of the spadix itself in length. Nevertheless this 
beak is part of the spathe, which in this group is highly specialized, and very 
different from that of the Calami, although, as in the latter, all principal ramifica- 
tions of the spadix, or partial inflorescences, have their own proper spathe. 
Such spathes are not in the least tubular, but are concave and of the shape of 
an elongated fusiform boat. Of these primary spathes, the outermost is the largest, 
and includes within it all the others, inclusive of the beaks belonging to each. 
The greater number are double-keeled at the back. 
When ‘the panicle within the spathes is ready to expand into flower, the outermost 
spathe splits open all down one side, and each of the inner spathes follows the 
example in turn. The outermost spathe of the Cymbospathae is invariably, though to 
a greater or lesser degree, furnished with spines more or less laminar, often subulate 
or aculeate or again even like bristles, but always straight, never uncinate or 
clawed. Of the inner spathes, the second and third usually bear a few dorsal spines, 
or they may be quite bare. 
The unopened spathes of the species belonging to Piptospatha are always of a 
more or less elongated and cylindrical form, and in that state are rarely thicker than a 
finger; from the first their spathes are all tubular, inclusive of the outermost, which, 
in these species, does not completely enclose the internal spathes: the latter issue from. 
its open apex in gradation like the tubes of a telescope, hence each internal spathe : 
rises more or less, according to its species, above the one below. 
The primary spathe of D. macropterus and of D. lamprolepis and also that of 
D. Motleyi form almost a connecting link with those of Cymbospatha, in so far as before 
the anthesis their external spathe does, except at the extreme apex, enclose the 
internal spathes completely, and only fails to entirely envelope them because it does 
not possess a beak capable of covering the tips of all the others. 
In the species belonging to the group of D. Draco, the spathes are thick and 
coriaceous, the external spathe frequently persists longer than the others, and is 
armed with short stout thorns, which are however never clawed. 
In other species of the group of the Piptospathae (D. cristatus, D. geniculatus, 
D. longipes, D. longispathus) the spathes are thin, paper-like, and' either bare or 
variously setulose. 
In no ease do the spathes of Daemonorops after the anthesis appear long and 
tubular, nor do they closely sheathe the axial portion of the spadix. In Daemonorops 
Ann. Roy. Bor. GARD, Carcurra, Vor. XII. 
