Ixxi 
the other being free, a new differentiation must arise ; for the two 
ends being very differently situated, the intermediate ones will 
also differ accordingly as they are nearer one end or the other. 
Here there is a cause for the differentiation of united individuals 
that does not exist in any branched or other symmetrical 
arrangement than a linear one. Some of the Salpidee show such 
a rudimentary linear aggregation, but their mouths and vents 
being lateral the individuals are so similarly situated that no 
differentiation need occur. A little consideration will show us 
that this is one of those cases in which perfectly transitional forms 
are not be expected. A permanent union of individuals in a 
linear series, such as to necessitate differentiation of function 
among them, could only be effected by a series of co-ordinated 
gradations, each of which would have so great an advantage 
over its predecessor as to necessitate its extinction in the struggle 
for existence. We cannot expect to find the union without the 
differentiation, or the differentiation without the complete union; 
and it will, therefore, be impossible to prove that such was the 
origin of any group of animals, except by showing that numerous 
- traces of separate individualities occur in their organization, and 
cannot be explained by any of the known laws of development or 
growth in animals not so compounded. 
In the structure of the lower Annelids we do find strong in- 
dications of such an ancestral fusion of distinct individuals. 
These animals are composed of segments, not merely superficial, 
but exhibiting throughout a wonderful identity of form and 
structure. Hach segment has its branchie, its enlargement of the 
alimentary canal, its contractile dilatation of the great blood- 
vessel, its ganglia, its branches from the nervous and vascular 
trunks, its organs of reproduction, its locomotive appendages, and, 
sometimes, even its pair of eyes. Thus every segment is a 
physiological whole, having all the organs essential to life and 
multiplication. Again, just as other compound animals increase 
by gemmation or fission, so do these. The embryo leaves the 
egg a globular ciliated gemmule; elongation and segmentation then 
take place, always inthe hinder part, so as to elongate the compound 
animal without interfering with the more specialized anterior 
segment. In the Nemertide, and some Planaria, spontaneous 
fission occurs, each part becoming a perfect animal, and in the 
Teenia this is the usual mode of reproduction. ‘The account given 
