566 The American Naturalist. [July, 
lines of adaptations which are not preexistent in the germ plasma. 
We find that new characters of the second class do follow such 
purposive or directive lines, arising simultaneously in all parts of 
the organism, and first appearing in such minute form that we 
have no reason to suppose that they can be acted upon by selec- 
tion. The old view of nature's choice between two single char- 
acters, one adaptive, the other not adaptive, must be abandoned, 
since the latter do not exist in the second class. 
Fourth. — The most serious obstacle to the Lamarckian priticiple 
is the problem of transmission. How can peripheral influences be 
transmitted in the way we have outlined — now that we have such 
strong evidence for the continuity of the germ plasma ? If ac- 
quired characters are not transmitted it is clear that the whole 
Lamarckian principle is undermined, and all these instances of 
sequence e;cpress no causal relationship. We are then, however, 
left without any adequate explanation of the laws of variations of 
the second class, and are thus driven to postulate some third, as 
yet unknown, factor in evolution to replace the Lamarckian 
T N presenting to the Association certain considerations regard- 
ing methods of teaching geography, I venture to assume 
that your interests in educational matters extend so far down as 
to reach a subject which many scholars " finish " early in their 
course, and whose advanced study hardly receives its due place 
in our colleges ; certainly it has suffered from neglect. My own 
practice in the way of teaching it has been with college students 
in the division of physical geography, and not feeling entirely 
satisfied with the system of study as presented in the text-books 
in current use, I have endeavored to discover and supply certain 
elements by which instruction in the subject might be advanced. 
» A lecture delivered before the Scientific Association of Johns Hopkins Un 
on February 13. 1889. 
