SYSTEMATIC POSITIOX OF MITSASTEirOX. 201 



■whetliev tlie said cbaracters neces.saril_T pi-event it fi-om Ijeing takeu into the 

 same family, or in other words whether the differences are such as separat« 

 geuus from genus or family from family. 



So far as I am aware, a discussion of this kind ou JHtrastemou lias 

 uever beeu nudertakeu. Aloreover, there is uo work on the anatomy of the 

 intramatiical tissue, the flower-pediuicle, or tlie sti-uctm-e of the ovules, 

 which collectively make up a group of cliaracters wlrich are as important as 

 any other floral charaeters for the determinatiou of a systematic positiou. 

 In this paj)er, I shall fii-st consider the morphology of the plaut both exterual 

 and intemal, especiall^* so far as is uecessary for the present descussion 

 and theu iu its Ijearings ou the systematic position of the plant itself. 



n. Extemal Morphology of Mitrastemon, especially with 

 regard to its systematic Position. 



^The plants aro leafless, unbranching, siugle flowered parasites growiug 

 ou exposed roots of Querciis cuspidata. Roots over 1 cm. iu diameter frequent- 

 ly liear numbers of these plauts. The volva is always iu colouy, separate, 

 or couflueut Ijelow, short, globose, erect, sessile, dark fulvous. Pedimcles 

 simple, thick, straight or sometimes slightly cm-ved, terete or subcompressed- 

 ly terete, smooth on surface, provided with scales. Scales opposite in 5 or 

 6 paii-s, imbricate, decussate, concave, margiu aduate to the peduncles at the 

 liase, forming tubes in the basal portious, where houey-like fluids are secret- 

 ed in abundauce.** Scales are more or less shiuing owing to a small quanti- 

 ty of a fluid secreted ou tlie surface, especially wlien the flowers are iu ljl(X)m. 

 Flowers ai^e always hermaphrodite aud usually regular, but sometimes more 

 or less diclinous, esiJecially iu the ealier stage of flower-developmeut. Of 

 the fom- primordial leaves which cjmpose the periauth, the front aud back 

 ones are usiially broader than the lateral pair. lu many cases, they, in a 



• The reader is requested to refer to the original description of Mitraslem<in Yamamotoi Mk. 

 contributed by Mr. T. Maein-o to the Tokyo Bot. Mag. XXV. p. 252, T. VII. .ind .1/. Kaicasa- 

 sakii Hat. in the same journal, XXVI. p. 112. 



•* In the flowering sea.son, thia honey attracts large numbers o£ a bird resembling a Japanese 

 white eye, {Zo.iterops japonica T. et Z.). 



