140 A. W. Bartlett. 
texture they are membranaceous and the aborted ovules can be 
discerned through them. 
The three flowers of the uppermost verticillaster (Fig. 16, E 
and F, and 17, A have the gynceceum most modified. One of these 
(Fig. 16, F) shows pronounced phyllody while in another one E we 
see a more advanced stage of that drawn in Fig. 16, B in that the 
ovary has become still more enlarged and the style correspondingly 
shorter. But the most curious feature in these two (Fig. 16, E and F) 
was that the floral axis in both flowers had produced a stalked flower- 
bud, which had grown up inside the ovary and the outline of which was 
just visible through the carpel wall as I have shown by the thin 
internal line in the drawings. In one of these flower-buds, that 
shown in F, the calyx and corolla could be distinctly made out 
by careful dissection, but only three epipetalous stamens with the 
typical diverging anthers could be found, and the centre was 
occupied by two leaf-like carpels quite separate from each other. 
In the other flower-bud (E) dissection was more difficult as the 
parts were somewhat distorted, the five calyx-teeth were not plain, 
there were four separate rudiments of petals and to three of them 
anthers were adhering; I could find no trace of the fourth stamen. 
The two carpels were separate and foliaceous, the apex of each 
was bent outwards representing a rudimentary stigma and each 
bore two rudimentary ovules. 
The gynoeceum of the third flower of the uppermost whorl of I. 
presented the much reduced form shown in Fig. 17, A, viz., a small 
rounded body, slightly exceeding 1 mm. in diameter carried on a 
short stalk and bearing at its apex three small erect pieces. The 
drying and subsequent boiling had obliterated the internal structure 
and I was unable to ascertain what it represented morphologically. 
In inflorescence II. the pistil had undergone much the same 
modifications as in I. In the two lowest verticillasters of flowers 
(B) it closely resembled that in the three lowest false whorls of 
I. In the third and fourth (Fig. 17, C and D) it consisted of two leaf¬ 
like carpels united by their edges for rather more than half their 
length. 
In the sixth false whorl (Fig. 19, E) the two carpellary leaves have 
become quite separate, they are unequal in size and show no trace 
of ovules, and between them at their base is a minute rudimentary 
flower-bud. 
Of the three flowers constituting the highest verticillaster, two 
contained each a short stalk, due to the elongation of the internode 
