258 Macfarlane. — Observations on 
than as the result of an exaggerated dorsal fusion of two 
originally lateral leaflets. Various weighty reasons may be 
urged in favour of this view. As we have already stated, the 
lid in some species arises as two lobe-like processes which 
ultimately fuse. In adult lids the free extremity may be 
more or less indented ; also two strong diverging vascular 
bundles come off from the recompacted midrib. Further, if we 
compare the supposed condition here with other leaf-fusions, 
we shall discover that it is not a peculiar one. Thus the fused 
or connate axillary stipules of various orders of plants originate 
exactly in the way that we imagine the lid of Nepenthes to do. 
Both are foliar parts which, growing round dorsally, fuse to 
form a median laminar process. Again, another parallel 
is seen in connate leaves, for if we suppose two connate 
leaves to develop entirely on the axillary or dorsal side of 
their supporting branch, i. e. dorsally instead of bilaterally, 
they will by fusion appear as a projecting leaf growing towards 
the main axis, and traversed by two midribs. But I think 
that a highly important confirmation of the above view is to 
be had in the presence of paired lateral or unpaired median 
lobes on the filiform midrib extremity beyond the point of 
origin of the lid. A feature which at first sight goes against 
the view just propounded is exhibited in pitchers of N. Veitchii , 
etc., where a strong ridge-like thickening occurs on the inner 
lid surface in the median line. But when examined micro- 
scopically this is seen to consist merely of cellular tissue, and 
is of no morphological importance. It need scarcely be added 
that the xylem of the lid bundles lies facing the pitcher cavity, 
while the phloem is directed upwards, though this is an 
additional confirmation of the above. 
We may fairly view a Nepenthes leaf then as being compound 
and composed of from 3 to 4 or 5 pairs of leaflets disposed 
along a midrib, and that both leaflets and midrib undergo 
striking modifications, the leaflets in particular showing a 
marked tendency to dorsal fusion above. In Plate XVII, Fig. 7, 
I have given an ideal diagram of an entire leaf of N. Rajah. 
First, the basal part of the midrib carries two elongated green 
