208 SYSTEMATIC POSITION OF MITHASTEVOl'. 



in a very few cases, very primitive ducts, annular or spiral, within the 

 tracheidal group. I could not discover any true well-formed ducts. Neither 

 sieve-tubes, nor companion cells, nor bast fibres, nor wood-fibres are found 

 anywhere. There are one or two collenchymatous rather rounded rectangular 

 cells near the bundles. The whole tissue like the host itself contains a 

 considerable amount of tannin. The structure of the peduncles is, upon the 

 whole, quite similar to that of Cytinm, differing from it only in that it 

 lacks well-formed pitted or annular ducts. 



IV. Systematic Position of Mitrcistemoti. 



As is stated above, Mitro.^temon bears a close affinity to plants at 

 present referred to the RarHesiacese and especially to those of the A pod a n- 

 theae, a tribe of the same family. It also resembles in some respects the 

 Nepenthacea>, Hydonoracese, Aristolochiaceae and Balanophora- 

 ceae (MAKING II., p. 253). It comes near the Nepenthacefe (WuNSCH- 

 MANN) in having single perianth and superior ovary; but greatly differs 

 from it in having peculiar parasitic organs, united stamens, one-celled ovary 

 with many parietal placentas and gamophyllous perianth. It resembles the 

 Hydonoracese (SoLMS-LAUBAOH T.) in its parasitism, its leafless, one- 

 flowered-peduncles, but differs in having united stamens, one-celled ovary 

 with many parietal placentas and berry-form fruit. Its affinity to the 

 Aristolochiacena (SOLEREDER) is to be seen in the gamophyllous perianth, 

 one-celled ovary, berry-form fruit and short columnar styles ; but it is quite 

 distinct in its parasitism, leafless peduncles and united stamens. It some- 

 what resembles the Balanophoraceae (ENGLER) in having a parasitic 

 habit, and scaly leafless flower-stalk ; but the difference in floral structure is 

 so great that it hardly needs pointing out. 



As we have seen in external as well as the internal morphological 

 characters, Mitrastemon has a close relation to the RafHesiacepe, in its 

 absorption-organ, uniflowered leafless peduncles, tetramerous perianth, one- 

 celled ovary with parietal placentas, short columnar style and stigma, and 

 in the structure of its ovules. The only points which formerly led us to 



