884 THE JOURNAL OF BOTANY 
Going back to the flowers, the male of Juliania has a perianth ; 
the female, none. In Corylus the conditions are reversed; in 
Betula, neither sex has an obvious perianth ; in Quercus, the flowers 
of both sexes are furnished with a’ perianth. 
of the Cupulifere have an ovary which is more than one- 
Weighing the characters in which there is agreement or 
similarity between the Julianiacee and the Anacardiacea, and those 
in which there is agreement or similarity between the Julianiacea 
and the Cupulifere, the latter in my estimation preponderate ; an 
cannot suggest a more natural position for the Julianiacea, in 
a linear arrangement, than between the Juglandacee and the 
Cupulifere. . 
BIBLIOGRAPHICAL NOTES. 
XL.—Artas per Draromaceen-Kunpx. HERAUSGEGEBEN voy A. 
‘ Scumuipr. 
Since some uncertainty attaches to the dates of issue, especially 
_ of the earlier parts, the following notes may be of service to those 
consulting the work. 
It has been issued in Hefte, each containing four plates. With 
the first 20 hft. a single folio of Vorléufige Evléuterungen was issued 
i d 12. 
hft. 1 and 4, are mentioned. This preliminary text was super- 
seded in the first 20 hft. (pls. 1-80) by Eritéuterungen, consisting 
of a single folio for each plate, that were issued with a second 
edition of the plates. This fresh text also gives what purports to 
be the dates of first issue, but these are in some cases manifestly 
erroneous. 
This Revidirter Text was also published without the plates 
for the benefit of former subscribers (Nature Novitates, May, 1887, 
p- 121). 
With hft. 21 (pl. 81) began the custom, still continued, 
. issuing a single folio of Vorléufige Erliuterwngen with each 
plate. 
Two indexes to the names of the species (Verzeichniss, &c.) have 
been issued. One to plates 1-144 (Series I-III) in 1890, and a 
second to plates 1-240 (Series I-V) in 1902, 
