218 Experimental Zovlogy 
based on these results. He lays much emphasis on what he 
calls physiological units or unit-characters, but disclaims any 
desire to locate these in any special part of the cell, and believes 
the evidence justifies him in supposing that each new step or 
mutation involves a change in the organization of the germ-cell of 
such a kind that a new unit-character appears. The change 
may be slight or great compared with the parent form, but its 
unity is shown by its behavior in heredity. New elementary 
species are characterized by having at least one new physiological 
unit. By way of example I may cite the following cases of 
mutations amongst animals. The appearance of the merino 
ram was probably due to some physiological change in a germ- 
cell of one of its parents of such a sort that one character espe- 
cially, viz. the wool, was changed. Other less striking characters 
were also present, and inasmuch as these are always associated 
with the merino type of hair, they all belong together. The 
ancon ram showed differences in the proportion of nearly the 
whole body; but the entire change must probably be referred 
to a single, although profound, unit change in the germ- 
cells. The occasional appearance of the turnspit type of dog, 
which resembles the ancon ram in the form of its body, indicates, 
if it does not prove, that this kind of change, involving nearly 
all the parts of the body, may readily occur. So numerous are 
the parts affected in these cases that it is impossible to ascribe 
the results to ‘‘accidental” combinations of the different organ 
elements, for it is inconceivable that just these combinations 
could ever appear more than once, but must be due to some 
definite change in the germ-cells which can appear in this 
particular way. The same statement must also hold for 
Lamarck’s primrose, where, however, instead of one possible 
combination there are several possible ones that become realized. 
If we use Galton’s oft-quoted comparison of a polyhedron, we can 
see that resting on its most stable face it may be rocked back and 
forth, but always returns to its same resting-place. Such oscil- 
lations on a single face would correspond to the fluctuating varia- 
tions of a species. Should a greater movement take place, the 
