25 



AFFINITIES. 



Bentliam, B.P1. iii, 193, places it in a subsection with E. piriformis, lo gi- 

 folia, urnigera, and ccesia, contrasting it with E. longifolia. 



He also points out that E. decuroa, E. doratoxylon, and a few others have 

 loose reflexed umbels. (See the figures of the umbels of buds of E. leucoxylon, 

 Pig. 13.?, Plate 55; E. decuroa, Pig. In, Plate 70; and E. doratoxylon, Pig. 4d, 

 Plate 70, of the present work.) 



Mueller contents himself with, "The leaves remind one of those of E. 

 amygdalina, though they are smaller ; the lid is much like that of E. iereticomis ; 

 the fruit is not verv similar to that of any other convener." I do not think that 

 comparisons with these two species can be usefully considered. 



Xor am I impressed with the affinities suggested by Bentham in the first 

 paragraph. E. erythronema has a good deal of individuality, and at present I am 

 inclined to look for its close relations amongst the Cornutce. I will return to the 

 subject when that group of plants is being dealt with. 



