Chapter VII —97— Byblis 



is sufficiently deep and narrow so that the sides of the glands lean 

 against and are supported by the sides of the furrow. 



The stalked gland (14 — 12-15) has a capital of usually 32 cells 

 radiating from the centre and standing out like an umbrella top. These 

 cells all abut on a central short cell resting on the top of the long stalk 

 cell. This in turn stands on a group of basal cells which may be as 

 many as eight in number, or as few as two in the case of a small 

 stalked gland. The latter may also have as few as four cells in the 

 capital, the mature glands showing no great degree of uniformity in 

 this regard. The stalk cells of the larger glands have strongly striated 

 thick cellulose walls, the striations reaching deeply, as far as the cuticle. 

 These striations run obliquely (as in the cotton fiber) and when the 

 gland dries (in air or alcohol), the stalk cells twist, as noted by Dar- 

 win {14 — 14). Fenner regards this arrangement as one to allow 

 bending of the trichome without collapse. 



While the gland capitals are covered with a thin cuticle there is 

 access by diffusion through pores, mentioned but not described by 

 Fenner. I found them (19336) to be rather large oval openings ar- 

 ranged in a circle about and some distance away from the centre of the 

 capital. They become evident on treatment of the stalked glands with 

 sulfuric acid {14 — 12). Both the sessile and stalked glands are 

 readily penetrable by dyes (methylene blue). 



Our earlier knowledge of the function of the glands bearing on the 

 question of the carnivorous habit of the plant we have at the hands of 

 A. NiNiAN Bruce (1905). Her work is clearly indicative of this, but 

 the question needs further investigation, which in this type of plant is 

 not easy. Bruce placed minute cubes of coagulated egg-albumen in 

 contact with the sessile glands, and after a period of some days (two 

 to eight) the whiteness has completely disappeared. During the 

 progress of digestion the round white core of the cube of albumen could 

 be observed to suffer gradual reduction in size. This material placed 

 in contact with the heads of stalked glands failed to show any evi- 

 dence of digestion, but when removed and placed in contact with the 

 sessile glands promptly did so. This seems to indicate that bacterial 

 action does not supervene. Some observations by Fenner justify 

 Bruce's results. When insects are caught, he says, and come in contact 

 with the sessile glands, a secretion is thrown out by them which is 

 much less viscous than that of the stalked glands. After four to six 

 hours, the group of glands affected again become dry and an examina- 

 tion of them shows that the contents of the gland cells and even of 

 the stalk cells betray evidence of absorption in the presence of a greater 

 density of the protoplasm and the presence of large rounded dark 

 masses. These changes are not to be observed in the stalked glands, 

 which do nothing else than secrete mucilage. I attempted to prove 

 the matter for myself at Perth, W. AustraHa. Byblis appeared late in 

 the season, during the latter part of my visit, so that I had only lim- 

 ited time at my disposal. My method consisted in placing minute 

 fragments of carmine fibrin in contact with the glands of the living 

 leaf, on the plant, and in a small vial with a dozen short pieces of leaf 

 with and without a little added water, with and without added weak 

 HCl, and with and without ammonium nitrate. The results were en- 



