J. M. Macfarlane. — Gephalotaceae. \ \ 



explained his mode of Separation of azerin, an active principle that he considered to 

 be present in pitcher liquids, and which caused rapid wetting of insect prey, in the 

 manner first described by Mellichamp for Sarracenia. In this connection Hamilton 

 says "there is no doubt that the liquid contains azerin or some similar principle, as 

 the insects falling into the fluid were immediately wetted through and drowned". 

 Accurate investigations are still needed to settle such points satisfactorily. 



A biological feature that again suggests parallelisms of evolution in Cephalotus 

 ■and other pitcher plants, is the presence in the former of an insect that seems normally 

 to pass its larval stage in the liquid of the pitcher. The writer has stated (Pflzreich. IV, 

 110p. \ 7) that for Sarracenia several species of insect so behave, that an insect 

 ■and a spider are intimately associated with Darlingtonia, while for Nepenthes the case 

 may be recalled of a spider which Everett describes as living within, and spinning its 

 web across, the pitcher cavity in order to catch attracted insects. For Cephalotus 

 Hamilton says "while watching some plants in their native habitat, we noticed flies 

 hovering around the pitchers and occasionally entering them. One of these I captured. 

 Mr. Froggatt informs me that it is one of the Tabanidae. It had the appearance 

 and blood-sueking habits of the ordinary March-fly of N. S. Wales. It is possible that 

 the larvae found in the liquid are those of this insect, and that the individuals we 

 saw entering it were intent upon depositing their eggs in the mass of digesting or 

 decomposing insects inside. In any case, the living larvae are an <jxample of one of 

 those cases, not of symbiosis, but of one organism taking advantage of the conditions 

 created by another for its own benefit." 



Floral structure and pollination (Blütenverhältnisse und Bestäubungseinrich- 

 tungen). The Flower-stalk or rhachis (Fig. 1 A) begins to appear in November, and 

 by mid-December may have reached a length of 30 — 50 cm but opening of the flowers 

 occurs in January. During the lengthening stage 7 — 10 narrow deciduous sterile 

 bracteoles are scattered along the axis. The lowest of these are dose to the normal 

 leaves, and they may be 1,5 — 2 cm long. The upper and also similar ones that subtend 

 the individual branches of the inflorescence are clustered and incurved. Schweig er 's 

 statement therefore "Vorblätter wurden an den Blüten nicht beobachtet" has evidently 

 been made from study of mature shoots. They are finely pilose, and amongst the 

 hairs are numerous glands. By the time that the rhachis has fully lengthened most 

 or all of the bracts have fallen. The upper part of the rhachis bears 7 — 12 short 

 scorpioid cymes, that arise each in the axil of a fertile deciduous bract. Fach cyme 

 bears 3 — 8 flowers. At the time of first blooming the lowermost 1 — 3 cymes are 

 loosely separate from the upper, which form a clustered mass, but in the later blooming 

 and in the fruiting stages most or all of the cymes may be markedly separate and 

 expanded. The stronger pedicels and even the feebler secondary ones — that are alike 

 ilevoid of bracteoles — are abundantly covered with thickened. hairs, with stomata and 

 with glands. 



The six sepals (Fig. 1 C) are slightly connate, and are of a purplish white color. 

 Their exterior surfaces are richly provided with stomata, also with hairs that are rela- 

 tively shorter, stronger and more sharply warted externally than any other set. The 

 inner surface is destitute of hairs and stomata. Each is somewliat hooded at the tip, 

 and is inserted on the edge of the expanded receptacle. 



The twelve stamens are developed as two circles of six each (Fig. 1 B). The 

 outer six are slightly longer and stronger than the inner, while they alternate with 

 them and with the sepals. They also dehisce earlier. All are perigynously inserte<l 

 on the expanded receptacle. Tlie filaments are subulate, slightly incurved and purplish- 

 white to pink. The anthers are subglobose, purplish white, and they dehisce introrsely. 

 From the outer side of the connective a subglobular cellular mass is formed (Fig. 4 A, B) 

 that is eminently characteristic and was selected by Labillard iere for origin of the 

 generic name. The yellow pollen grains closely resemble those of Crassulaeeae, Saxi- 



