1891.] Are Acquired Variations Inherited ? 211 
to the existence of causal relationship ; this probability would be 
increased if it could be shown that no other explanation of this 
class of variations will stand the same test. 
First, as to sequence. The overwhelming majority of variations 
as observed in the fossil series” occur along the lines of use and 
disuse. Weismann has urged that all variations in this class are 
substantially quantitative, that where an organ becomes stronger 
by exercise it must possess a certain degree of importance, and 
when this is the case it becomes subject to improvement by 
natural selection. It follows from embryological development 
and the laws of growth by cell division that all new characters 
are in one sense quantitative, but in tooth evolution we have 
examples of the rise of structures which are qualitative,—v.¢., 
essentially new, and not simple modifications of ` preéxisting 
forms. I refer to the successive addition of new cusps. As 
already observed, there is absolutely no evidence for indefinite 
variation in these characters. The new cusps do not rise spon- 
taneously at random points and then disappear, to be replaced by 
the gradual development of those which happen to rise at adap- 
tive points.* One of the most surprising recent discoveries is 
that one after another these successive cusps are added to the 
simple conical crown at the point of maximum wear ; that is, the 
most-worn points in an earlier series of generations are those at 
which the new cusps appear in the later series. 
Paleontologists cannot, however, claim that this sequence is 
universal. Among the rare exceptions there are, first, some sec- 
ondary cusps® which arise from the base of the crown,—~. £., 
entirely out of the region of use and disuse and pursue the same 
steady development until they reach a stage in which they are 
obviously useful and subserve attrition. Second, upon the principle 
that the action and reaction of two opposing surfaces must be 
equal, it is difficult to explain some cases in which we observe a 
the ee studies of Kowalevsky, Cope, and Ryder among the vertebrates, and 
of Hyatt, Dall, and others among the invertebrates. 
54 Biol. Memoirs, p- 84. 
85 See “ The Evolution of Mammalian Molars to and from the Tritubercular Type,” 
AMERICAN NATURALIST, December, 1888. 
86 Such as appear in some molars of the later Tertiary ungulates. 
