596 FRAGMENTS OF SCIENCE. 
Combining duly the two factors, all the previous irregu- 
larities disappeared, every result obtained receiving the 
fullest explanation. On studying the account of this 
masterly investigation, the words wherewith Pasteur him- 
self feelingly alludes to the difficulties and dangers of the 
experimenter's art came home to rne with especial force: 
" J'ai tant de fois eprouve que dans cet art difficile de 
1'experimentation les plus habiles brouchent a chaque pas, 
et que 1'interpretation des faits n'est pas moms peril- 
leuse." * 
CHAPTER XXXVI. 
SCIENCE AND MAN.f 
A MAGNET attracts iron; but when we analyze the 
effect we learn that the metal is not only attracted but 
repelled, the final sipproach to the magnei being due to 
the difference of two unequal and opposing forces. Social 
progress is for the most part typified by this duplex or 
polar action. As a general rule, every advance is balanced 
by a partial retreat, every amelioration is associated more 
or less with deterioration. No great mechanical improve- 
ment, for example, is introduced for the benefit of society 
at large that does not bear hardly upon individuals. 
Science, like other things, is subject to the operation of 
this polar law, what is good for it under one aspect being 
bad for it under another. 
Science demands above all things personal concentration. 
Its home is the study of the mathematician, the quiet 
laboratory of the experimenter, and the cabinet of the 
meditative observer of nature. Different atmospheres are 
required by the man of science, as such, and the man of 
action. Thus the facilities of social and international 
intercourse, the railway, the telegraph, and the post office, 
which are such undoubted boons to the man of action, 
react to some extent injuriously on the man of science. 
Their tendency is to break up that concentrativeness which, 
as I have said, is an absolute necessity to the scientific 
investigator. 
* " Comptes-Rendus," Ixxxiii., p. 177. 
f Presidential Address, delivered before the Birmingham and Mid- 
land Institute, October 1, 1877; with additions. 
