380 Agassiz on the Relations between. Animals 
At this stage of the investigation, a comparison between Mol- 
lusca and Radiata shows that, though the former advance farther 
in their fluviatile development, and even reach with some few of 
their types a terrestrial mode of existence, there is not yet a single 
family among them which is entirely terrestrial, nor a single class 
which is either entirely fluviatile or terrestrial, this connection 
with the higher conditions of existence being only introduced 
among some few of their representatives, which we are allowed 
from other data to consider as the highest in their respective 
groups. 
If we now pass to the great group of Articulata and begin as 
before by revising their zoological arrangements as based upon 
anatomical and embryonic data, we shall have at the outset to 
settle the limits of their classes, and their relative positions. 
The first point which we have here to investigate is the ques- 
tion whether Articulata in the widest extension of this group con- 
stitute one single natural type, or whether they should be subdi- 
vided into two equivalent groups, as has been proposed by those 
who would restore the division of worms, in its widest sense, as a 
great division equal in zoological importance to the type of Mol- 
lusca, and unite the Arthropoda, Crustacea, and Insects to form 
another group of equal value. 
The great diversity among worms, seems at first to warrant in 
some degree, such an arrangement. But as soon as we consider 
the metamorphosis which insects undergo, and compare their 
earliest stages of growth with the structure and forms of worms, 
we cannot fail to perceive, that notwithstanding the many pecu- 
liarities which characterize worms, they are, after all, only one of 
the permanent modifications of the same type as Crustacea and 
insects, among which last the characters and forms of a large num- 
ber of worms are reproduced as transient states of growth ; so 
that upon the most natural view, and especially if we allow embry- 
ology to have its due weight in fixing our opinion, we must con- 
sider worms with all their diversified forms, Crustacea in all their 
diversity, and Lepades, Arachnid and Insects, to constitute one 
single undivided natural type in the animal kingdom. Assuming 
upon the foundation alluded to, and without entering into a de- 
tailed argument upon this question, that this is the right view of 
this subject, the next question is about the number of classes into 
which these Articulata should be subdivided. T'aking here again 
anatomical and embryological evidence as our guide, and remem- 
ring what was said above of intestinal worms we shall find that 
the most natural combination of the different groups of Articulata 
will bring them all into three classes, one containing those in 
which the body is either more or less distinetly articulated, or 0 
which indications of transverse wrinkles in the skin are scarcely 
marked, or wholly wanting, but in which, however developed these 
