ROTATORIA OF THE UNITED STATES. 
301 
pointed out of late, Gosse was quite mistaken upon this point; these species have 
two equal or unequal bristle-like toes. This genus Codopus must then of course be 
dropped, as a synonym of Diurella. Mastigocerca as used by Gosse is equivalent 
to Rattulus in its primitive application, so that it must be replaced by Rattulus. 
Finally, the group of species which Gosse distinguished as Rattidus can not well be 
separated from Diurella , and should be included in that genus. 
Lord (1891) founded a new genus, Elosa, said to belong to the Raftulidee, but 
lacking a foot. From the description which he gives of this animal, it seems clear 
that it is not one of the Rattulidaz, but belongs rather with Ascomorpha. 
Division of the Rattulidre into genera. — We may now inquire a little more fully 
into the basis for classification in this family. On what grounds can the Rattulidre 
be divided into genera? 
The characters which have been used by previous authors are mainly two: (1) 
the presence or absence of a lorica; (2) the number and relative length of the toes. 
As to the first point, Ehrenberg distinguished two genera, Monocerca , without 
a lorica, and Mastigocerca , with a lorica. In the former he placed Rattulus rattus 
Muller; in the latter, Rattulus carinatus Lamarck. Now, these two are so closely 
related that it is doubtful whether they should not be considered one and the same 
species, and both have the cuticula stiffened to form a lorica. The same is true of 
the other species, Rattidus longiseta Schrank, included by Ehrenberg in his non- 
loricate genus Monocerca. In fact, the distinction between Monocerca . and Masti- 
gocerca had no basis in reality, and it is quite impossible to divide the family in this 
manner, for all have a lorica. 
As to the second point, Bory de St. Vincent (1824) included in Monocerca the 
species having a single long toe, while Diurella had two evident toes. This distinc- 
tion, in one form or another, has been kept up and is in use at present. 
Tessin (1886) held that this was not a good basis for division into genera, for 
he saw in Rattulus gracilis Tessin (pi. v, figs. 45-49), a species which, with its 
shorter toe about one-tliird the length of the main toe, formed a transition from the 
single-toed to the two-toed forms. He therefore united all the Rattulidre in a single 
genus. 
There can be no question but that Tessin was right in believing that interme- 
diate stages could be found between the two-toed and one-toed forms. In fact, as I 
have shown in the account of the toes, almost every gradation can be found between 
the condition with two equal toes, and that where only a single toe can at first be 
detected, and all the species can be shown to have two toes, though the right one is 
in many cases a mere rudiment. 
If, therefore, we are to consider the genus a natural group, including only spe- 
cies that are more closely related to each other than to any species of another genus, 
I believe there is no escape from the necessity of classitying the Rattulidre all in one 
genus. I have made many attempts to group them into what seemed natural genera 
on other bases than the toes, but found that all had the same defects; some of the 
species within the genus were apparently not so closely related to each other as they 
were to some species outside the genus. 
Perhaps the nearest to a natural group within the family would lie made by 
separating off Rattidus capucinus Wierz. & Zach. , R. cylindricus Imliof, and R. 
multicrinis Ivellicott as a separate genus. But R. elongatus Gosse is very closely 
