280 Experimental Zoology 
distal end, but this is characterized by the presence of a grow- 
ing region, which continues to add new segments to the new 
part. 
What factor determines that the éerminal organs are those that 
are first laid down in the new part? It cannot be the shape 
of the new part as such, for this is practically a dome-shaped 
knob for all new parts. The bounding surface seems certainly 
to be a factor in this relation of the parts, as does also the relation 
of the old organs or layers at the cut surface. Between these two 
boundaries the relation of the parts to each other determines the 
result. A number of considerations, that I cannot enter into 
more fully here, have led me to suspect that this relation of the 
parts can be accounted for as due to a condition of stratifi- 
cation or polarity, due to the mutual pressure of the parts on 
each other, which acts as the stimulus for the differentiation of 
the cells. By these same assumptions we can, I think, also give 
a fairly consistent explanation of the difference in the rate of 
growth at different levels. 
Let us take, by way of illustration, the results that have been 
obtained in another worm, lumbriculus. If the worm is cut in 
two at almost any level, there develops from the posterior end of 
the anterior piece a new tail, and from the anterior end of the pos- 
terior piece anew head. The material out of which these two new 
parts develop must be identical. What determines, then, that 
the new material forms in one case a head and in the other a tail ? 
Since the development of these new parts seems to be largely a 
centripetal phenomenon, we cannot assume that the influence of 
the old part on the new, a centrifugal influence, determines the 
result; but since the order or sequence of the differentiation in the 
new part is the same as that in the old part, this may determine 
whether a head or a tail develops. In other words, the polarity 
of the new part is in each case the same as that of the old. This 
polarity is an expression of the stratification of the differentiation; 
at least, this is the most probable view of polarity, I think, that 
we can find at present. The centripetal influence acting on 
the new material at the anterior end determines therefore that 
