640 British Association . [September, 
the scientific foresight and the industry which were required to 
frame hypotheses that were worth testing, to guide the investiga- 
tion by these hypotheses, to contrive, construd, and adjust 
adequate apparatus, and to make with it the elaborate observa- 
tions and the exad observations and maps which were necessary. 
But when by these means the new fads had been brought to 
light, the inference from them that there is iron in the atmo- 
sphere of the sun was an easy one. This example will better 
convey than a definition what are the charaderistic features of 
an experimental inquiry. 
On the other hand, no series of observations or experiments, 
however skilfully arranged, could have enabled anyone to under- 
stand the cause of that familar but truly surprising phenomenon 
that a top stands upon its peg while it is spinning. But a full 
explanation of it is within the reach of any student who will 
train his mind to reason consecutively, and avail himself of the 
aids to prolonged consecutive thought which mathematicians 
have contrived. He will then see that the most obvious and 
familiar mechanical fads involve as necessary consequences all 
the phenomena which he finds in the schoolboy’s top, in the 
physicist’s gyroscope, and in the precision and nutation of the 
heavens. This, then, is a problem of Nature which falls within 
the province of the dedudive method. 
Whatever data are known exadly, these inferences from these 
data, however remote, may be depended upon as corresponding 
with what adually occurs in Nature. And if in such cases the 
mind of man has proved equal to the task of drawing inferences 
which can effedually grapple with the problems he finds around 
him in the Universe — which is, alas ! as yet but too seldom — 
then will the dedudive method, our plummet, explore depths in 
the great ocean of existence which our anchors of experiment 
could not have reached. 
We must bear in mind that either method of investigation 
may be misapplied, and that this is a risk carefully to be guarded 
against. The dedudive method when misapplied lands us in 
speculation ; the experimental method becomes empiricism ; and 
it so happens that the sciences of mechanics and chemistry are 
not only monuments of the power of the two great methods of 
investigation, but instrudive examples of their weakness also. 
For in chemistry scarce any attempt at prolonged reasoning, 
carrying us by any lengthened flight to a distance from the 
experiments, can be relied on. The result has seldom risen to 
anything better than speculation. And, on the other hand, 
in mechanics, conclusions which depend on experiments only are 
empirical ; that is, they are deficient in accuracy, and their rela- 
tion to the other phenomena of the science is left in darkness. , 
Here, then, we find in these two sciences not only how strong j 
these two methods of investigation are, but how Weak they may | 
become if misapplied. 
