138 







in so far as the chalazogamy itself or its transition to the porogaray fonud in 

 the two series is concerned !) ; and also in all probability to the Ericaceee in 

 the anatomical characters" 5 . Its position should, therefore, be shifted from 

 one place to another, according to each one of its natural relations. 



Next, follows the Urticales' . The affinity of this series to the Verticillatse, 

 Juglandales and Fagales has been pointed out before. 



Here also ENGLER puts the Proteales ; as to this series, our knowledge 

 is as yet very limited ; the only relation that is thus far known to us is to 

 the Santalales, as the Proteacere and Loranthaceffi show some agreement in 

 the perianth -lobes and in the andrceceuin. 



Next, come the Santalales' according to ENGLEK'S system. He places tin's 

 series here in a fixed position, but its rather complicated relation to other series 

 clearly indicates the mutability of its natural position according to the view 

 taken of it. Thus, the affinity of tins series to the Proteales, as has been 

 mentioned, is clear. Then, the resemblance borne by the Santalales to the 

 Coniferee or to the Gnetacese is so very striking that it not only presents 

 external analogy, but also suggests many morphological relations' . Moreover, 

 to some extent it approaches the Sapindales, as the Olacace?e 7) and Icaciiiacese 

 show a perfect similarity in their fruits and external features. Further, it 

 bears also some resemblance to the Kosales, as can be seen in the Grubbiacaae 

 and Hamamelidacefe which show agreement in their wood-anatomy. Certain 

 of the Santalales (Balanophoraceae) 9) are, by some authors, compared with the 

 Myrtiflorse from the view that Cynomortum (Balanoph.) and Hippmis (Halorrhag.) 

 show some resemblance in their bisexual flowers. Finally, the relationship of 

 the Santalales and Pthamuales is recognizable, when we compare the Loran- 

 thaceffi and Vitacefe, taking the calyculug of the Loranthacefe or Viscum for 

 a reduced calyx. Thus, the Sautalales is related to the Proteales, Couiferse, 

 Sapindales, Rosales, Myrtiflora3 and finally to the Rhamnales. Its natural 

 position is, therefore, changeable from one place to another, so that it should 



1) WETTSTEIN, E, E. 1. c. p. 591. 2) JEFFREY, E. C. 1. c. p. 385. 



3) ENGLEK, A. Erl mterangen, 1. c. p. 363. 4) Nat. Pa.-fam. III. 1, p. 176. 



5) ENGLEE, A. Erlaiiterangen, 1. c. p. 363. 6) Nat. Pfl.-fam. III. 1, p. 211. 



7) Nat. Pfl.-fam. III. 1, p. 233. 8) Nat. Pfl -fain. III. 1, p. 229. 



9) Nat. Pfl.-fam. III. 1, p. 2i9. 



