NODAL ANATOMY AND THE MORPHOLOGY OF STIPULES 449 
be homologous with the stipular appendages of the other mono- 
cotyledons which we have discussed. Anatomical evidence therefore 
points toward a connection between nodal topography and stipular 
appendages in the monocotyledons similar to that which we have 
noted in the dicotyledons. 
Let us now consider the various exceptions to the general rule that 
lateral leaf-traces and stipules always occur together. 
There are about 20 unilacunar families in which stipules are 
found, but in most of these families they are rare and are almost always 
thin, scarious or minute. If our theory is correct, these unilacunar 
families have been derived by reduction from trilacunar (and hence 
presumably stipulate) ones, and it would therefore be only natural that 
stipules or vestiges of them should occasionally persist. In many 
cases the vascular supply of a unilacunar leaf, although causing but a 
single gap, is composed of three bundles; and the stipules, when present, 
are often related to the lateral ones of this trio. The interpetiolar 
stipule of the Rubiaceae is evidently to be regarded as a fusion of 
two formerly independent and adjacent ones, since each stipule 
receives its vascular supply from both leaf-traces. 
There are also about 30 families which are prevailingly trilacunar 
but in which stipules are absent or rare. Some of the more im- 
portant of these are the Juglandaceae, Proteaceae, Menispermaceae, 
Aristolochiaceae, Lardizabalaceae, Calycanthaceae, Pittosporaceae, 
Simarubaceae, Burseraceae, Meliaceae, Aceraceae, Plantaginaceae 
and Cucurbitaceae.2 Many genera and species in normally stipulate, 
trilacunar families often lack stipules. The lateral bundles at the 
node in such plants do not cause the formation of stipules, but directly 
opposite each lateral trace is often observable a swelling or rounded 
projection which only needs to be slightly elongated, or to be the seat 
of a gland or water pore, to become a typical stipule. 
If we examine these exstipulate trilacunar families, as set forth 
in the previous table, we are at once struck by the fact that in the 
great majority of them (about 75 per cent) the leaves or leaflets are 
prevailingly entire. This suggests the possibility that stipules, on 
the one hand, and the teeth or lobes of the lamina, on the other, may 
be dependent for their occurrence on essentially the same factor or 
factors. 
Such an hypothesis is strengthened by a study of the relation 
2 The tendrils of the Cucurbitaceae are perhaps stipular in character. 
