972 THE AMERICAN NATURALIST. [Vor. XXXV. 
difference at the ends, and since this is found to be true for 
every smallest piece of the magnet, it is assumed to be a 
property of each molecule or atom of which the magnet is 
made up—the magnet being the sum total of the action of all 
its molecules. We speak also of a crystal showing polarity in 
the sense that it has a definite form that can be referred 
to definite poles and axes, and it is supposed that this same 
property is possessed by the smallest * crystal molecules " of 
which the crystal is made. 
On the other hand, in an animal or in a plant in which we find 
the two ends of a piece behaving differently, yet always show- 
ing the same constant difference, although we have a result 
resembling in many respects the polarity of the magnet, we find - 
also in several essential respects differences and, I think, fun- 
damental differences (in the sense that they are the outcome of 
different factors) between this phenomenon and that in the 
magnet. For instance, we find in the pieces of Tubularia that 
external conditions may reverse entirely, after a time, the 
polarity of the piece, so that a head may appear on the poste- 
rior end even while another head is present on the other end. 
We know of nothing similar in the case of the magnet, unless 
we consider the change that can be brought about in it by 
re-magnetization; but there are quite obvious differences in 
the two results. The most fundamental difference, however, 
between the two conceptions is, I believe, connected with the 
kind of structure that we must suppose to be present in the 
two cases. The polarization of the magnet is the sum total of 
a vast number of smaller units, each unit being itself a magnet. 
On the other hand, we have at present clear evidence to show 
that while the polarity of the organism is the outcome of its 
structure or organization, this is not the effect of the presence 
of a large number of units of which the organization is only 
the sum total, but, on the contrary, the organism is itself the 
ultimate unit that makes the polarity, and while we must 
believe that this same organization may develop in a piece of 
the original whole, the evidence goes to show that the reor 
ganization is of a different nature from that by which a 
piece of a magnet becomes af once a new magnet. 
