183 
remote relation is to be found between the Taccaces and the Iridaces in the 
form of the stigmata, on the one hand, and on the other, between the former 
family and the Dioscoreaces in the undivided ovary. Moreover, there exists 
some resemblance in the Taccacess and Burmanniacee. 
38 Dioscoreacee”: Very clearly related to the Liliaceze, especially to 
the Smilacoides of the latter family. Another connection is to be found 
between this family and the Taccacee or Amaryllidaces. 
39 Iridaceze”: These resemble the Heemodoracez in the single whorl of 
the androeceum, though the latter are represented by the outer whorl in the 
Tridacez, while by the inner whorl in the Heemodoracese. The Morxes of the 
Tridaceze approach the Taccaces in the shape of their stigmata. 
y 
Series X. Scitaminece 
40 Musaces”: The coincidence in the floral structure in the Musacem 
and Orchidaces is the most precise conceivable in the comparison of orchida- 
ceous flowers and Lowia and Orchidantha both of which show their lips 
directed downwards by resupination. 
41 Zingiberacezw: As above. 
42 Cannacee: As above. 
43 Marantacere”: These are nearest the Cannacee; while the latter 
are related to the Zingiberaces. The four families of the Scitamines are con- 
nected at one point or another. 
Series XI. Microsperimece 
44 Burmanniacer”: Some relations seem to exist between this family 
and the Orchidaces, and also between it and the Amaryllidacee. 
45 Orchidacee™: This family is near the Burmanniacee and the 
Triuridacee. It agrees to some extent with the Musacee, Zingiberaces 
and Marantaces in its imperfect staminal whorls. The diagram of Musa is 
in accordance with that of Arundina pentandra. 
1) Nat. Pf.-fam. IL—5, p. 132. 2) lo. IL—5, p. 141. 
3) ScHumann, K. O.—Musacee, in das Pflanzenreich, IV—45, pp. 11-12, et p. 41, Fig. 10. 
4) Nat. PA.-fam. IL-6, p. 38. 5) Le. IL-6, p. 46. 
6) Lv. IL—6, p. 75. 
