494 
On Scientific Method. 
[October, 
may most reasonably look for extensions of our knowledge : 
at all events we are warranted by the recent history of 
natural philosophy in so doing.”* 
As an illustration of the use made by genius of “ residual 
phenomena” I might cite the discovery of the planet 
Neptune by Adams and Le Verrier. Slight anomalies were 
observed in the motions of Uranus : these were studied ; the 
hypothesis was framed that the peculiar movements were 
due to the presence of an unknown body ; observations 
were carried out, and the new planet was discovered. 
Almost every science presents us with residual pheno- 
mena awaiting explanation. To mention one in chemical 
science. Why are the densities of the vapours of phosphorus 
and arsenic twice as great, and the densities of the vapours 
of mercury and cadmium one-half as great, as all analogical 
reasoning would lead us to imagine they should be ? Here 
is an unexplained facT which will doubtless one day be pro- 
lific of consequences. 
I have thus attempted to sketch the main points in that 
method which has been, and is, pursued by scientific men 
in their attempts to discover the truths of Nature: in con- 
clusion I must say a few words regarding the limits of 
scientific method. 
In science we start with fatfts, we then form hypotheses 
which we test by appeal to facTs. But so great is the num- 
ber of facTs presented to us that we cannot observe or 
experimentally determine more than a small, almost an 
infinitely small, portion of them. Much less can we hope to 
form satisfactory hypotheses which shall explain them all. 
This is true in physical science. The mass of facts gathered 
together by the naturalist is already extremely large ; but 
there can be no doubt that the number of the unknown 
vastly exceeds that of the known facts of Nature. And of 
the known facts how few have as yet been explained. The 
problem of the “ mutual effects of three bodies, each acting 
on the other under the simple hypothesis of the law of 
gravity,” can scarcely be said to be yet completely solved. 
And if this comparatively simple case has puzzled the in- 
genuity of the mathematicians what are we to say to the 
application of mathematical processes to the explanation of 
those motions and mutual acftions which we have reason to 
believe are being performed and undergone by the constituent 
portions of every chemical atom ? Each of these particles, 
Sir J. Herschel has remarked, is continually solving differ- 
* Thomson and Tait, The Oxford Pamphlet, p. 108, 
