1929] 
Four New Forms of Eriophyes 
297 
shall be given to a form that is similar to another but is 
limited to a different host species? Shall we regard it as 
a variety, a species, or what? 
The discussion of varieties Nalepa brushes aside with the 
remark : “Fur den prakticken Cecidologen ist die Kenntniss 
der Varietaten der Arten and Unterarten bedeutungslos.” 
According to his dictum one must distinguish between the 
fluctuating variations of a species within itself and the 
constant variations in a biological sense. Then : “Als Varie- 
taten werden wir grundsatzlich nur jene Variationen des 
Arttypus bezeichnen, die das gleiche Cecidium auf ver- 
schiedenen, jedoch verwandten Wirtspflanzenarten verur- 
sachen.” And . . erzeugen die Subspecies verschiedene 
Gallenformen auf ihrer gemeinsamen Wirtpflanze.” 
Cotte protests that much of Nalepa’s work is useless and 
that a system of taxonomy can be acceptable only if fixed by 
common consent, not by one person. He condemns the sub- 
jugation of morphological to biological characters practiced 
by Nalepa, a new host is not sufficient evidence to establish 
a new variety. The objections of Cotte are justified pro- 
vided Nalepa’s rule is followed without any further con- 
firming data. If distinct morphological characters are not 
forthcoming to distinguish the form in question, then con- 
trolled observations on the physiological limitation of this 
form to a certain host plant are necessary to justify estab- 
lishing it as a variety or subspecies of another similar or 
even identical morphological form on a closely related 
species of host. 
Where it has not been possible to control the infestation 
and so determine whether the mite forming an identical or 
similar gall is limited to the specific host and where the 
morphology of the mite itself is relatively close to a known 
species, but still different, after allowing for the variations 
within the species, I have considered it to be a variety of the 
species already described from a similar gall on a plant of 
the same host family. This is the case with the two new 
varieties presented in this paper. In the case of the one 
new subspecies, E. rudis dissimilis, the separation is based 
on differences of morphology in the mite and the gall form 
with which it is associated as well as its limitation to the 
