147 



Juglandales and Resales, all belonging to the Arclu'clilamydese, have been 

 alluded to above. It remains to compare it with the Matachlamydefe. As to 

 the Ebenales and the Tubiflorce, their relations to the present series are 

 manifest in the single coated ovules in the Limnauthacese , in the partial agree- 

 ment of the SalvadoraceBB' 5 with the Oleacece, and in that of- the Aquifoliacese^ 

 with ;the Symplocacese. As to the affinity of the Sapindales and the Ericales, 

 this is suggested in the agreement of the Cyrillacefe 4 ^ and the Ericaceae, in 

 the habit and in the peculiar anther of Costcea of the former family, and of 

 the Pentaphylacace88 6) and the Clethraeeae in the pored anthers and the capsules. 

 Further, some relation is to be found between the Sapindales and the Primulales, 

 as^can be seen in a comparison of the Corynocarpacege n) and the Myrsiuacese 

 (G. DON, ENDLICHER). Finally, their relation to the Coniferoe, as in the case 

 of the Geraniales, is shown by the presence of the very similar resin - ducts 

 in the Anacardiacese 7) and the Pinaceae. As can be seen from the statements 

 above given, the groups so far considered are not to lie regarded as serially 

 related, but are in close intermixed relations. To arrange them in one order 

 or another is quite right in this respect or that. One can not have a claim 

 superior to another. Natural positions for the two series are found in several 

 places between the Conifers and the Metachlamydere. 



Next, come the Rhanmales 8) . As to this series, he says : 



Die Rhamnales smd jetzt anf die tetracyklischen Archichlamydeen mit vor den 

 Bib. stehenden Stb. beschrankt. Da bei den Rhamnaceoe die Stellung der Raphe an 

 der auf steigenden Sa. sehr wecliselnd ist, so konnen die Vitacecc, welclie immer ventrale 

 Raphe liaben, Tinbedenklicli neben die Ehamnaeeai gestellt werden. 



Its relations to the Resales and Geraniales have already been referred to. 

 Its relation to the Umbelliflorae is, in my estimation, undeniable. A comparison 

 of the Vitaceae 9) with the Araliacese and the Umbelliferfe will at once justify 

 this conclusion. One might object to the above statement on the ground that 

 the Rhamnales have stamens opposite the petals, while the Umbelliflorse have 

 stamens alternate to the petals. But this, in my opinion, makes no great 



1) WETTSTEIN, R E. 1. c. p. 017. 2) Nat, Pfl.-fam. IV. 2, p. 19. 



3) According to Dr. R. KANEHIEA, both families are very similar in anatomical characters. 



4) Nat. Pfl.-fam. HI. 5, p. 180. 5) Nat. Pfl.-fam. Nacht. I. p. 215. 

 6) 1. c. Nacht T. p. 217. 7) 1. c. III. 4. p. 234. (Burseraceas). 

 8) ENGLER, A. ErHxitenongen, 1. c. p. 367. 9) Nat, Pfl.-fam. III. 8, p. 111. 



