84 
THE JOURNAL OF BOTANY 
name .... when the group which it designates embraces elements 
altogether incoherent, or when it becomes a permanent source of 
confusion or error.” This should be interpreted by the examples 
supplied. Elements altogether incoherent. —“ The genus Schre- 
hera L. derives its characters from the two genera Cuscuta and 
Myrica (parasite and host) and must he dropped; and the same 
applies to Lemairea De Vr., which is made up of elements belonging 
to different families.” The case of Cleyera is not comparable. The 
type-species of Cleyera comprised species of two genera, hut both 
belonged to Theacece. The generic description applied exclusively to 
one of them, and was sufficiently characteristic to enable I)e Candolle, 
and Siebold and Zuccarini, to identify the genus. The specific 
description combined characters of C. japonica and Ternstroemia 
gymnanthera, but it is possible to exclude the latter from Cleyera on 
account of its non-agreement with the generic description. No one 
has proposed to set aside JBignonia L., although it comprised thirteen 
genera of Bignoniacece and one each of Loganiacece and Vitaceoe ; 
nor to treat J Bignonia indica L. as embracing elements altogether 
incoherent, although it included two species belonging to different 
genera. A permanent source of confusion or error. —“ Linnaeus 
described under the name of Bosa villosa a plant which had been 
referred to several different species, and of which certain identifica¬ 
tion seems impossible. To avoid the confusion which results from 
the use of the name Bosa villosa, it is preferable in this case, as in 
other analogous cases, to abandon the name altogether.” But there 
is no doubt as to the identity of the genus Cleyera and of the species 
C. japonica , and no confusion has resulted from the use of these 
names. 
The name Ternstroemia japonica Thunb. was a new combination 
for Cleyera japonica , and should be treated as a synonym of that 
species. As applied by Siebold and Zuccarini to the Ternstroemia 
erroneously included by Thunberg in C. japonica, it has proved to be 
“ a permanent source of confusion or error,” and should therefore be 
superseded by T. gymnanthera (Wight et Arn.) under Art. 51, 4. 
Eiioteum and Freziera. 
The type-species of Eroteum is undoubtedly E. theceoides, as 
stated by Fawcett and Bendle. The type-method determines the 
incidence of a generic name once and for all: the generic name is 
always associated with the type-species. Hence if Eroteum theceoides 
is included in Cleyera , the genus Eroteum automatically becomes a 
synonym. Under the method of residue adopted by Fawcett and 
ltendle, the name Eroteum may be applied to two different genera, 
according to circumstances. “ If it is agreed to pass over the original 
Cleyera Thunb., then Eroteum Sw. is the earliest name for the 
genus now generally known as Cleyera. If Cleyera is to be retained, 
then Eroteum Sw. must he used for the genus of which E. undu- 
latum Sw. was the first described species; that is, the genus now 
