I 



AFFINITIES OF BALANOPHORE^. 27 



Trans, xx. p. 103); who however erroneously refers Bhojialocnemis {Fhceocordylis, Griff.), 

 of which his specimens appear to have been imperfect, to the Moiwsti/li. The genera of 

 each group are all more nearly related to one another than to any of those of the other 

 group ; and such a di\asion is therefore perfectly natural : but there are, notwithstanding, 

 such very gi'eat differences between the members of each group, that the genera of 3Ioiio- 

 styli especially have all characters of far more than generic value, and may well be 

 conceived to be types of very distinct assemblages of genera. This is not so much the 

 case among the Distyli, for the two genera of Lophophytea; are very nearly related, and 

 those of the Ilelosidece so mvich so, that it may be doubted how far Coryncea, Scybaliimi, 

 and Spharorkkon are distinct from Uelosis. 



Sarcophyte must at present be considered as in many poijits a doubtful member of the 

 Monostyli, from my inability to discover whether the ovarium is simple or compoimd, at 

 any period of growth. The absolutely sessUe, discoid stigma is found neither in JIoiio- 

 slyli nor Distyli, and in some other characters it partakes as much of one group as of 

 the other ; thus in its three stamens it agrees with Balanophora and all the Ilelosidece, 

 but it differs from them in its stamens being free, and in its anomalous anthers (which 

 however sviggest an affinity with B. polyandra). In its branched inflorescence it resembles 

 Lophophytea;, but differs totally in the structtire of the male flowers. 



For characters of secondary importance, I have availed myself of the male flowers, and 

 especially of the nmnber and form of the stamens and their cohesion. A remarkable 

 analogy between two of the sections into which each primary groiip is divided, is esta- 

 blished by these organs ; the Lophophytece [oi Distyli) having free stamens of the ordinary 

 type of Phsenogamic plants, and being hence analogous to Cynomorium and Mystrope- 

 talon (among Monostyli), while aU the other genera have anomalous anthers. 



In framing the genera, characters of the third degree of importance have been employed ; 

 such as the presence or absence of the perianth, and its structure ; aided occasionally by 

 the stamina, the cohesion of the flowers, the nature of the inflorescence, and lastly, the 

 habit ; this term implying in these plants not merely differences in the outline of organs, 

 but also in their development and mode of evolution. 



Besides Griffith's arrangement oi Bcdanophorew, that of Endlicher (Meletemata, p. 4) is 

 the only one of any importance hitherto proposed. His division is founded on the stamens 

 being free or combined : this however not only assumes the position and arrangement of 

 the stamina to be of more importance than those of the ovaria, but brings together genera 

 which have otherwise little in common. 



Geographical Distribution, and Variation. 



The greater number of Balanophorem belong to the tropical and subtropical mountains 

 of Asia and South America, where they probably occur in nearly equal proportions. In 

 both hemispheres certain species ascend to 10,000 feet ; comparatively few being found in 

 low tropical forests, almost the only ones being Balanophora fungosa in the eastern, 

 Selosis guyanensis in the western hemisphere, and Thonningia in Aft-ica. A considerable 



e2 



