SYSTEMATIC POSITION OP MITRASTEMON. 207 



tubes caused by so-called " obliteration " (Tschirch p. 345); wbile tbe 

 secondary bast of tbe latter is not at all swoUen, and there is present a 

 considerable amount of tbe depressed lamellae of sieve tubes. We aLso find 

 here that tbe swolHug of the bast-tissue is especially lai-ge in the floral 

 polster. This is possibly ascribable to the same cause as that to which we 

 ascribe the similar effect iii the case of gaUs (Kekner p. 472) and mycor- 

 rhiza (Shibata 1. c.) 



The inti-amati-ical tissue, whcn fully grown, produces endogeueously a 

 flower-bud, which breaks up the bark of the host-root, and protrudes up- 

 wards a peduncle, just as in the case of a plant of the Kafflesiacese. (SoLiis- 

 Laubach I., p. 3 ; n., IV., Lotsy p. 833). 



2. Sti'uctiu'e of the Peduncle. 



The peduucle of Jlitrmtemon, as is usuaUy the case with a parasite, is 

 extremely primitive in its anatomical structure. The tissue is formed by 

 nearly round parenchymatous ceUs which are largest towai'ds the center of 

 the tissue mass and become graduaUy smaUer towards the epiderm. Inter- 

 ceUuIar spaces are greater towards the center and smaUer towards the 

 periphery, aud there are practically uone at all outside of the ring of vascu- 

 lar bimdles. The epiderm has no stoma. There are 5-10 horse-shoe shaped 

 or semi-circular, outward-cmTing, Tascular bundles around the central laith. 

 They are scattering when few in number, but come into close contact, when 

 thei-e are mauy of them. Towards the epiderm ou the outside of this 

 broken riug of vascalar bundles, there are sometimes seen a very few smaU 

 bundles which seem to run into scales on the pedimcles. There is generaUy 

 in the center of a buudle, in cross section, a group of a few spiral ti-acheids 

 aiTauged iu a cm-ved liue, which is almost completely suiTouuded by layers 

 of plasmatic cells. The layer of plasmatic ceUs is thicker on the outer side 

 of the ti'acheidal bundles, than on the iuner side. The cambium-layers, or 

 rather a few cambium-ceUs, are fouud just outside of, and close to, the 

 ti-acheidaI-gi"oups. The ti-acheidal ceUs here represent a primitive xylem, 

 and the sun-oundiug plasmatic cells an as yet undifferentiated bast. I found. 



