420 
CARL S. HOAR 
ovules. He considered that the two families have been developed 
independently, the Fagaceae being the more primitive of the two (a 
view not held by all). Hallier (1901 and 1908) attempted to demon- 
strate the affinity of the Cupuliferae and the higher Angiosperms. 
In doing so he considered the Cupuliferae as derived from the Hamame- 
Hdaceae and through these from the Laurineae. In this way he would 
connect them with the Rosalean forms allied to the Combretaceae. 
It is noteworthy in this connection that in 1908 he discarded the view 
held in his earlier paper and considered the above as derived from the 
Anacardiaceae and Burseraceae chiefly because of the strong anatom- 
ical likeness of Juliana to those orders upon the one hand and to 
Juglans upon the other. Clearly the observer had very small proof 
upon which to base his belief and hence, perhaps, his reason for chang- 
ing his opinion. Goebel (1905), through his work, came to the con- 
clusion that the Fagales and Juglandales are reduced forms. The 
superior gamophyllous perianth, syncarpous ovary, and complicated 
inflorescence, he thought not characteristic of a primitive family. 
Arber and Parkin (1907) also took issue with those who consider the 
same cohorts as not being reduced forms. Recently Berridge (1914) 
observed that the Amentiferae are not an isolated group but have an 
obv"ous relationship with the higher Angiosperms. In her article, 
the inflorescence, flowers and cupule, of Castanopsis chrysophylla are 
described and the anatomical structure of the flower is fully worked out. 
This species is compared to Castanea vulgaris, Fagus sylvatica, Quercus 
Robur, and Juglans regia. She found the flower differing in no essen- 
tial from the epigynous types of the angiospermous flowers and drew 
a comparison between the Rosaceae and the CupuHferae, since it 
seemed probable to her that the epigynous Rosaceae, or their near 
descendants, are the forms with closest affinity to the ancestors of the 
Fagaceae. 
Among those who hold a different view upon the subject we find 
such men as Treub and Engler. Since those who believe in the 
reduction theory have worked freely with the Juglandales, it is well 
to note the work of other investigators upon the same group. Nawas- 
chin and Finn (1913) worked with Juglans regia L. and Juglans nigra 
L. and came to the following conclusions: Among seed plants there is 
the tendency to reduce the male gametophytes from sperms to naked 
nuclei. Together with this reduction occurs the evolution of the 
pollen tube. The species of Juglans studied have binucleate genera- 
