J. M. Macfarlane. — Cephalotaceae. 13 



Epidermiszellen wölben sich papillenartig empor. Der obere Teil des kreiseiförmigen 

 Höckers verbreitert sich nach der Peripherie hin und damit werden die einzelnen Balg- 

 früchte nach außen hingedruckt. Durch die dadurch entstehende Spannung werden 

 die Zellen am Grunde der Einzelfrüchte zunächst in der Nähe der kreiseiförmigen An- 

 schwellung zerrissen; die Zerreißung der Zellen am Grunde nimmt in demselben Maße 

 zu, wie die Anschwellung sich vergrößert, und schließlich fallen die so völlig aus dem 

 Verbände mit dem Blütenboden gelösten einzelnen Früchte aus der Blüte aus.« 



The enclosed seed is attached to the base of the fruit cavity, and shows mem- 

 branous testa and tegmen. It is of a broadly elliptic outline and encloses (Fig. \ F) 

 a dense granulöse albumen, at whose base the embryo is embedded. The ernbryo is 

 minute, straight, with short radicle and elongated cotyledons. 



Geographica! Distribution (Geographische Verbreitung). From the time of 

 Labillardiere onward for about 100 years botanists recorded Cephalotus only from 

 a limited area of South Western Australia. Ils discoverer Labillardiere found it at 

 Esperance Bay, a few years afterward B. Brown gathered it — as have others 

 frequently since — at King George's Sound, Drummond gathered it at Swan River, 

 Pritzel at S. W. Plantagenet, while Andrews gives "Swamps, Albany" the region 

 where also Hamilton and Di eis found it. Di eis (op. cit. p. 357) states the distri- 

 bution to be "Deep River to Esperance Bay". From settlers Statements and the above 

 records it seems likely that the plant occupies a coastal strip of about 500 km. How 

 far inland it may extend we do not as yet know. In a letter to the writer Di eis 

 kindly described as follows. "The whole region is one of the most temperate on the 

 continent of Australia. The underlying rock is granite. The soil is sand, blackish from 

 humus, and fairly wet from a more or less constant supply of water slowly soaking 

 its way to the surface." Hamilton's Statements exactly agree with this, while alike 

 the soil and its moist condition exactly remind one of Sarraccnia localities. Diels 

 further added, "the accompanying plants are small Restiaceae, Xerotes, Schizaea, 

 Dracophyllum and other Epacridaeeae, Conospermum nudiusculum, mostly a dense 

 growth of many-stemmed delicate slender species of about the same height as the 

 scapes of Cephalotus itself". 



It was first introduced into cultivation in 182 3 by Captain King, who collected 

 it at King George's Sound and brought it to the Royal Gardens at Kew. There "it 

 flowered repeatedly and ripened seeds from which several plants" were raised. Later — 

 and in recent years frequent — consignments have made it familiär in many greenhouses, 

 but its successful cultivation requires attention. It thrives best when grown like the 

 Sarracenioids in a pot of light humus-sand, the lower part of the pot being kept 

 constantly in water to a depth of about 2 cm. It does best with greenhouse treat- 

 ment, and does not prosper when exposed to continuous hothouse temperatures. 



Systematic Relations (Verwandtschaftliche Beziehungen). As indicated by the 



taxonomic quotations for the family at the beginning of this memoir, CepJtalotus has 

 been variously viewed by botanists. Relying largely on its perigynous and apocarpous 

 details Labillardiere — and at first R. Brown — placed it in Rosaceae. The 

 latter after study of its seeds strongly advocated the creation of a monospecific family, 

 the Cephaloteae, that showed affinities with Rosaceae, Saxifragaceae, Cunonieae and 

 Crassulaceae. Jussieu, according to Brown, had already viewed it as an aberrant 

 genus of Crassulaceae. Lindley also recognised the distinctness of the type, but in 

 spite of its perigyny inclined to view it and Dionaea as types related to Ranunculaceae. 

 Bentham included it as a very distinct form of the Saxifragaceae, and not a few 

 botanists have accepted this. Engler followed Brown in keeping the family name 

 and rank. In view of Dickson's andEichler's investigations, as well as the floral 

 morphology, the writer regarded it in 1889 as of family rank, and not related to 

 Sarraceniaceae- Nepenthaceae, though showing many analogous histological details in 



