10 J. M. Macfarlane. — Nepenthaceae. 



also it results from growth and recurving of a circular area below the rim, that ap- 

 pears as a circular swelling below and outside of the lid in the young pitcher. The 

 zone of tissue just beneath the recurved portion often develops a copious feit of hairs 

 that attain their maximum in N. albo-marginata, the rim of which is subtended by a 

 white tomentose band. The convex surface of the mature rim is firm and shining, 

 owing to heavy cutieularization of its epidermal cells. It is traversed by fine or pro- 

 minent parallel ridges that are above and in line with vascular bundles. Its surface 

 affords a very insecure foot-hold for insects. The ridges are usually prolonged beyond 

 the infolded margin as teeth of varying length. While the ridges and teeth are very 

 feebly developed in N. Reinivardtiana, they become long evident processes in N. mel- 

 ampliora, N. villosa and N: Edwardsiana. But their most striking development is 

 reached in N. eehinostoma, where the entire rim has become split up into a double 

 row of incurved teeth that project toward the pitcher mouth. Round the infolded 

 margin of the peristome, or toward the tips of the teeth are pores that correspond to 

 those already mentioned in N. Loivii. These open into longer or shorter canals, that 

 are in line with, and act as ducts for, the deeply embedded marginal glands. From 

 these glands a sweet juice is discharged that is greatly relished by insects. In relation 

 to the size of the pitcher, the widest and most perfectly inclined peristome is that of 

 N. ampullaria (Fig. \ 3). Those of JV. Northiana, N. rajah, N. sanguinea and N. 

 Veitchii form a wide conspicuous often richly colored "frill" round the mouth. But 

 more commonly the peristome is a nearly uniform circular collar, as in N. gracilis, 

 N. khasiana and N. phyllamphora. It attains huge proportions in N. villosa (Fig. 9) 

 and JV. Edwardsiana (Fig. 16, p. 53) where it is a broad cylinder elevated at rather 

 distant intervals into plate-like ridges that are continuous with the long marginal 

 teeth. In several species, notably N. Hemsleyana and N. Rafflesiana the peristome 

 and pitcher wall are greatly elevated posteriorly into a high neck that bears the lid. 

 The upper halves of the peristome, in the latter species, are widened out, while the 

 ridges and teeth are decurved toward the pitcher orifice. This becomes more pro- 

 nounced in N. bicalcarata, where the uppermost parts of the peristome are lengthened 

 out, decurved, and tapered into two hard sharp pointed spurs or spines that overhang 

 the pitcher orifice (Fig. \ i). Burbidge's explanation of their significance seems good. 

 He observed in North Borneo that the pitchers of many species are visited by the small 

 rodent Tarsius spectrum. Perched on the pitcher margin, it bends in its head and 

 neck, scoops out the caught insects and devours them. But if it attempts such action 

 with N. bicalcarata the two sharp spines often transfix it by the nape of the neck, 

 and tumble it into the pitcher, or frighten it from attempting such action on other 

 pitchers of the species. Another suggested explanation of the spines has recently been 

 made, by supposing that they exude honey drops by their tips from a few marginal 

 glands that are so placed as to cause an insect that attempts to sip, to drop off into 

 the pitcher cavity. Such may be a partial reason for their gradual evolutionary selec- 

 tion and development, but Burbidge's view seems more natural. 



The lid or operculum (Fig. 4e) is always developed, but it varies from a small 

 narrow elliptic process as in N. ampullaria (Fig. 1 3) to large cordate or reniform 

 expansions as in N. rajah and N. bicalcarata (Fig. 14). Bower and the writer have 

 both suggested, from embryological (Fig. 3 c) and morphological evidence that the lid 

 is to be viewed as two peltately fused laminar lobes or leaflets. Bower regards them 

 as leaflets that are distinct in morphological value and continuity from the laminar 

 lobes in front of the pitcher, and from the basal laminar halves. The writer views 

 them as lobes of the primitively continuous lamina, that have early become rounded 

 off and fused in peltate fasbion in front of the spur to form the lid, while later and 

 less perfect Separation of the pitcher wings and the laminar lobes has occurred. Eacb 

 lid shows, in most species, two strong and somewhat approximated veins running into 

 it, and distributing minor median and lateral veins. Its apex is often emarginate or 

 indented. Ils external surface closely agrees with that of the outer pitcher wall, alike 



