132 The American Naturalist. [February, 
small number of metameres that become differentiated. Other 
features which are common to most Branchiates, but which 
either are not common to all or are at the same time common 
to some of other groups, will appear below. It must, of course, — 
be understood that in the above diagnosis of the group, fea- 
tures of internal anatomy are known only of recent forms; of 
the visceral structure of the trilobites, we are absolutely in the 
dark. 
CLAss I—CRUSTACEA. 
Branchiate Arthropoda with functional gills; with one or 
two pairs of distinctly preoral appendages (antennæ) the first 
being purely sensory; the ganglia corresponding to these 
appendages being fused with the protocerebrum to form the 
“brain ;” the appendages with typically a basal joint giving 
rise to two or three branches; several pairs of appendages 
modified for eating; alimentary canal with long cesophagus 
and well-developed stomodeal “stomach,” mid-gut region 
short, the mid-gut glands (“liver”) being well-developed; 
proctodeum long. 
In all living Crustacea (Eucrustacea), there are two pairs of 
antenne, although in some forms (e. g., Apus, Oniscids) one or 
the other pair has has become greatly reduced. In the Trilo- 
bites, on the other hand, but a single pair has, as yet, been dis- 
covered. It therefore remains to be shown whether a single ` 
pair is characteristic of these forms, or whether we have here 
a possibly greatly reduced additional pair. In case the former 
alternative prove true, it may be necessary to remove the 
Trilobites completely from the position here assigned them, 
though it will not necessarily follow that they should be aBa 
ciated with the Eurypterids and Limulus. (For the position 
of the Trilobites, see below). 
It is difficult to say exactly what weight should be given the 
so-called “typical Crustacean limb,” the di- or trichotomous 
appendage so frequently met with in this class, and which 5 
not infrequently regarded as diagnostic. That this condition 
is a derivation of the lamellate condition found in Apus, 
as maintained by Lankester (’81) admits of little doubt, and 
