MAGAZINE OE SCIENCE AND ART. 
207 
published three volumes of nautical information, ob¬ 
tained from Dutch vessels in the Atlantic and Indian 
Oceans. For the purposes of meteorological science 
this system cannot be considered as complete until 
observations oil land are included. Most of the greater 
atmospheric changes are due to the distribution of land 
and water and to the different effects of the sun’s rays 
on each. Observation alone can furnish the data from 
■which the effects of these agencies can ho calculated; 
and we can therefore probably make no great advance 
in the knowledge of the meteorology of the globe with¬ 
out a concurrent investigation of its two leading de¬ 
partments. Land observations exist in great numbers. 
In Prussia, in Russia, in Austria and in Belgium, such 
observations are organised under Government direction, 
or at least with Government support. What is needed 
is to give unity to these isolated labours — to 
connect them with one another, and with the results 
obtained at sea ; and the first step to this seems to be 
to give them, in each country, that permanence and 
uniformity of system which can only ho insured in 
measures adopted by the State- Here, however, we 
encounter an objection, upon which it is necessary to 
say a few words. It has been objected to the science 
of meteorology, as it has keen usually studied, that it 
proceeds upon a false method, and that consequently 
it has led, and can lead, to no results. I feel myself in 
a manner compelled to notice this grave objection : in 
the first place, because it proceeds from men whose 
opinions on this (or almost any other scientific question) 
are entitled to the highest deference; and secondly, 
because this Association -must bear no inconsiderable 
measure of the reproach, if it he well founded. First, 
then, as to the results: l am free to admit that the 
number of those who arc engaged in the discussion of 
meteorological observations is dispropOTtionnhlg smalt. , 
and that the results obtained probably fall far sboit of 
what may be expected from the data already accumu¬ 
lated. But that the methods have led, and can lead, 
to no results, is, I think, sufficiently:disproved by the 
labours of a single man—Prof. Dove, of Berlin. And 
if it be true that the. course pursued in the science has 
yielded much fruit, in proportion to the labor besto wed 
on the discussion, it will hardly be deemed widely 
erroneous. Still it is quite possible that the methods 
pursued, though not fruitless may be inadequate , and 
it seems therefore necessarily to notice the objection 
somewhat more minutely. It is asserted, then, that 
the capital vice of the science of meteorology, as at 
present pursued, is that it has no definite aim ; that it 
ought to embrace an enquiry into the physical consti¬ 
tution of the objects with which the science is concerned, 
and an investigation of causes as well as laws of 
phenomena. It may bo admitted, at once, that the 
physical constitution of the bodies whose changes wc 
are investigating is a proper object of study to the phy¬ 
sicist ; but it does not follow that it should necessarily 
he conducted by thu same individuals who are in search 
for the laws of the phenomena, or even that the former 
knowledge is essential to tho progress of the latter. The 
noblest of all the physical sciences—Asfro nomy —little 
more than a science of Laws —laws, too, of the simplest 
kind of change ; and the knowledge of these laws is 
wholly independent of the physical constitution of 
the masses whose movements it studies. A similar 
observation may be made regarding the science of Ter¬ 
restrial Magnetism ; and the case is one which brings' 
us still nearer to the question at issue, inasmuch as tfie 
laws have been obtained—and they are numerous— 
have resulted from a method of inquiry altogether 
similar to that adopted in meteorology. Time will not 
permit me to inquire whether there is not a miscon¬ 
ception of a metaphysical kind at the root of this 
objection. I may observe, however, before leaving the 
subject, that there 3Te two modes of studying the 
sequences of natural phenomena,—one in their relation 
to time, and which is best accompl ished by observations 
at stated periods, and the other in the relation of the 
'successive phases of tho phenomenon to one another. 
Of these, the latter, although not wholly neglected, has 
not been so much followed as it deserves; and I cannot 
but think that it would, if more systematically followed 
enrich the science of meteorology with s new harvest 
of results. 
The most important of the recent additions to the 
theory of Light have been those made by M. Jawin. It 
has been long known that metals differed from trans¬ 
parent bodies, in their action on light, in tills, that plane- 
polarized light reflected from their surfaces became 
elliptical!?/polarized ; and the phenomenon is explained 
on the principles of the wave-theory, by the assumption 
that the vibration of the ether undergoes a change of 
phase at tho instant of reflexion, the amount of which 
is dependent on its direction and on the angle of inci¬ 
dence. Tliis supposed distinction, however, was soon 
found not to be absolute. Mr. Airoy showed that dia¬ 
mond reflected light in a manner similar to metals; 
and Mr. Dale and Prof. Powell extended the property 
to all bodies haring a high refractive power. But it 
was not until lately that M. Jamin proved that there is 
no distinction in this respect between transparent and 
metallic bodies; that all bodies transform^ plane- 
polarized into clliptically-polarized light, and impress 
a change of phase at the moment of reflexion. Prof. 
Baughton has followed up the researches of M. Jamin, 
and established the existence of circularly-polarized 
light hv reflection from transparent surfaces. The 
theoretical investigations connected with this subject 
afford a remarkable illustration of one of those impedi¬ 
ments to the progress of Natural Philosophy which 
Bacon has put in the Ibrmost place among his examples 
of the Idola —I mean tho tendency of the human mind 
to suppose a greater simplicity and uniformity in nature 
than exists there. The phenomena of polarization com¬ 
pel us to admit that the sensible luminous vibrations 
are transversal or in the plane of the wave itself; and 
it was naturally supposed by Fresnel, and after him by 
M‘Cullagh and Neumann, either that no normal vibra¬ 
tions were propagated, or that, if they were, they had 
no relation to tiio phenomena of light. lYe now learn 
that it is by them that the phase is modified in the act 
of reflexion: and that, consequently, no dynamical 
theory which neglects them, or sets them aside, can 
be complete. Attention has been lately re-called to a 
fundamental position of tho wave-theory of light, 
respecting which opposite assumptions have been made. 
The vibrations of a polarized ray aTe all parallel to a 
fixed direction in tho plane of tho wave; but that 
direction may b& either parallel perpendicular to the 
plane of polarization. In the original theory of Fresnel, 
the latter was assumed to be tho fact; and this assump¬ 
tion Fresnel has been followed by Cauchy, In the 
modified theories of M‘Cullagh and Neumann, on the 
other hand, the vibrations are supposed to be paralel 
to the plane of polarization. This opposition of the two 
theories was compensated, as respects the results, by 
other differences in their hypothetical principles ; and 
both of them led to conclusions which observation has 
verified. There seemed, therefore, to be no means left 
to the theorist to decide between these conflicting hypo¬ 
theses until Prof. Stokes recently, in applying’ the 
dynamical theory of light to other classes of phenomena, 
found, one in which the eflects‘should differ on the two 
assumptions. When light is transmitted through a fine 
grating, it is turned aside, or diffracted, according to 
laws which the wave-theory has explained. Now, Prof. 
Stokes has shown that, when the incident light is polar¬ 
ized the place, of vibration of the diffracted ray must 
differ from that of the incident, the two planes being 
connected by a very simple relation, It only remained, 
therefore, for observation to determine whether tho 
planes of 'polarization of the incident and refracted rays 
were similarly related, or not. The experiment was 
undertaken by Prof. Stokes himself, and he has inferred 
from it that the original hypotheses of Fresnel is tho 
trne one. But as an opposite result has been obtained 
by M. Boltzmann, on repeating the experiment, the 
question must be regarded as still undertermined. 
r 
