BEOOM-RAPES, BALANOPHORE^, RAFFLESIACE^. 199 



Balanophoreffi in consideration of their coalescence with the roots of their hosts, 

 only three species are known. Two of them (Hydnora Africana and H. triceps) 

 belong to South Africa, the third (Hydnora Americana = Prosopanche Burmeisteri) 

 to South Brazil. The tuber is represented by a prismatic body with from four to 

 six angles furnished with papillae along the edges. The flower -buds which burst 

 from it have at first the form of spherical Gasteromycetes, but gradually elongate 

 and assume the form of a large fig or upright club. This structure opens at the 

 thickened upper extremity by three stout fleshy valves representing petals. At 

 the base of this curious flower no appendage is to be seen that could be interpreted 

 as a bract or leaf. The fleshy mass of flowers evolves a disagreeable putrid odour, 

 and in this property the Hydnorese resemble the Rafflesise, which belong to the 

 next gi-oup of parasitic Phanerogams. 



The fifth series of flowering parasites is composed of the RafilesiaceEe, plants 

 connected with Balanophoreas and Hydnorese by their general aspect, the absence 

 of chlorophyll, and the undiiFerentiated embryo which consists merely of a group of 

 cells. They used all to be classed together under the name of RhizantheEe; but the 

 Rafflesiaceae are now treated as a separate family on account of the characteristic 

 structure of their flowers and fruit. The formation of these organs will again come 

 up for discussion later on when we treat of the wonderful structure of the famous 

 giant-flower Rafflesia; at present we are only concerned with the relationship of 

 the parasite to the food-providing host-plant. This is, if possible, even more 

 remarkable than in the case of Balanophorese and Hydnorese. In the latter the 

 union is efiected within a structure like a tuber or a rhizome, the vessels and cells 

 of the parasite coalescing with the exfoliated and disordered wood-cells belonging 

 to the root or stem of the host-plant; whereas in RaiflesiaceEe the embryo, having 

 penetrated beneath the cortex of the host, produces a more or less definite hollow 

 cylinder which surrounds the wood of the host's root or stem (as the case may be), 

 and constitutes a sort of vestment intercalated between the wood and the cortex of 

 the host. There is no production of tuberous enlargements as in the Balanophorese. 

 The stem or root attacked by the parasite only exhibits a moderate thickening at 

 the place where the parasite dwells beneath the cortex, and the cortex itself is only 

 destroyed at the spot where the embryo pierces through it, and where subsequently 

 the flowers emerge. When roots constitute the substratum whereupon the parasite 

 has established itself, they are always of a kind that run throughout upon the 

 surface of the ground; when stems are chosen for attack, they are either the 

 branches of trees or shrubs, shoots clothed with dead foliage belonging to dwarf 

 sufiruticose bushes, or else woody lianes of tropical forests. The seeds are con- 

 veyed to the host-plants through the intervention of animals. 



Rafflesias are found in the haunts of elephants and along the tracks followed by 

 those beasts. The Rafflesia-fruits are accordingly no doubt trampled upon and 

 crushed, and the little seeds imbedded in the pulpy mass of the fruit thus have an 

 opportunity of adhering to the elephants' feet. The seeds are afterwards rubbed oft 

 by projecting roots at places more or less remote from the original locality, and if 



