J. M. Macfarlane. — Nepenthaceae. 7 



always well developed and of great tensile strength. Thus the tendril of N. Raffle- 

 siana may be loaded with a weight of six kilos without rupture. Like other tendrils 

 they are sensitive to contact Stimulus when young, and may then coil round supports 

 by two or three spirals, after which it is difficult to displace them. Only in short 

 upright or in creeping species such as N. rajah and N. Veitchii do the tendrils remain 

 straight and uncoiled. They are rarely uniform in thickness, usually they enlarge 

 suddenly or gradually as they near the pitcher base. An interesting biological condi- 

 tion of the tendril is seen in N. bicalcarata, the structural relation of which was first 

 explained by Beccari and Burbidge. Like the tendrils of most Nepenthes this has 

 alluring honey glands, but here they are exceptionally large and numerous, appearing 

 as evident papillae over the surface. Opposite the middle or lower half of the pitcher, 

 where these glands are most numerous, a swelling that is about twice the thickness 

 of the tendril above or below, is seen in wild specimens, and a similar though more 

 reduced swelling is seen in cultivated specimens (Fig. i 4). The swollen area is hollow- 

 ed out and occupied as a formicary by ants. About the middle of the area is a 

 circular hole, the entrance of the formicary (Fig. 1 4). The hole seems, at least in 

 some cases, to have been the duct of a honey gland that the ants have gnawed into 

 and enlarged while sipping its nectar. The liquid contents of the pitcher cavity, 

 filtering upward through the cells of the tendril, can be sipped by the insects in safety. 

 This constant liquid supply has caused enlargment and watery hypertrophy of the area. 

 Its central cells break down or are destroyed by the ants, and the cavity thus formed 

 becomes a formicary. As Beccari has shown it may vary in size from 2 — 6 cm X 

 I cm. From the standpoint of plant heredity it is significant to note that the swelling 

 always persists in cultivated plants, though these are not frequented by, nor perforated 

 by, appropriate ants. 



The next area of leaf specialization is [d) the pitcher or ascidium. The con- 

 stituent parts of this have been very variously interpreted by different botanists. Their 

 views are in part set forth by Bower and the writer [Ann. of Bot. III. (1839) 239, 

 254; IV. (1890) 165; VII. (189 3) 450]. The adult pitchers in some species are all 

 uniform or monomorphic, in other species dimorphic types of pitcher are produced, 

 while several are trimorphic in that three distinct forms of pitcher develop at different 

 heights on the plant. Alike from evolutionary, embryological and morphological evi- 

 dence, the primitive type seems to be the tubulär pitcher, that approximates in shape 

 to pitchers of some species of Sarraceniaceae. This shape is retained in most of the 

 monomorphic species such as N. gracilis (Fig. 2), N. khasiana, and N. phyllamphora. 

 But possibly through weight of the liquid contents of the lower or terrestrial pitchers 

 while these rested on the ground, some species bear ampulliform or goblet-shaped 

 pitchers below, while the upper suspended or cauline ones are tubulär or tubulär 

 ventricose (Fig. 1 5). N. alata, N. ampullaria, N. anamensis, N. melamphora and 

 others show such dimorphic conditions. But in N. Boschiana, N. maxima, N. Vieil- 

 lardii and less strikingly in a few other species, the lowest pitchers are ampulli- 

 form, those above are tubulär, while toward the upper part infundibuliform or cornu- 

 copioid pitchers are alone produced. Gradation types between all of these are usually 

 noted. 



As study of the seedling pitcher in relation to the adult pitcher shows (Fig. 3A — Z)j, 

 the often wide and almost flat area that intervenes between the wings of the pitcher 

 on the ventral surface, ' is the expanded upper or ventral side of the midrib. The 

 remainder of the pitcher wall is the inflated lower or dorsal side of the midrib. As 

 is proved by successive stages in the evolution of seedling leaves, the pitcher wings 

 represent the upper lobes of the originally continuous lamina, that become completely 

 separated from the inferior or basal halves, by intercalary growth of the midrib region 

 that becomes the tendril. These wings are usually well developed as lateral often 

 ciliated expansions that traverse the length of the pitcher in the terrestrial type of the 

 dimorphic species cited above. Their often richly ciliated margins are in striking contrast 



