of Chromatic Dispersion. 365 
tions in the wave-lengths which are rendered manifest by the 
displacement of the fixed lines from their normal relative positions 
in the unrefracted spectrum, and which are attendant on the 
irrationality. The first, or constant, portion of the index indi- 
cates the state of compression of the ether in the pores of the 
medium. The variable fragment consists of two portions,—one 
depending for its magnitude on the specific action of the pon- 
derable molecules of the medium on the ether within its pores, 
the other apparently depending on a specific action of the vibra- 
tions of the ponderable molecules of the medium on certain defi- 
nite waves passing through the ether in its pores, in virtue of 
which some of these waves are slightly lengthened, and others 
slightly shortened, but always in a certain symmetrical manner. 
To place the matter in a popular poimt of view:—Let the 
wether be regarded as a universally diffused elastic fluid, and 
each portion of it as comprising numerous centres of elastic or 
repulsive force. Since force cannot exist as a simple abstraction 
(for we cannot conceive of force as being exerted by nothing, or 
by mere space), it is needful to suppose these centres to be each 
occupied by a something which exerts the force, and which may 
be called ‘an cethereal particle.’ These particles must be sepa- 
rated from each other by certain minute but variable distances, 
otherwise there could be no movement of the particles. Each 
particle must be conceived to occupy a certain normal position 
in absolute space, from which it never departs except under the 
influence of some applied external force, and to which it always 
returns when that applied force ceases to act upon it. In the 
free ether, each particle is retained in its normal position with 
a certain persistence, in virtue of the forces exerted upon it by all 
the other ethereal particles by which it is surrounded. The force 
required to induce any given amount of motion in an ethereal 
particle must be proportional to the degree of this persistence. 
Assume, for the sake of illustration, that in the wave corre- 
sponding to the fixed line B in the free ether there are em- 
braced in the direction of its length 1,000,000 of such par- 
ticles, that is, the moving force has time to progress onwards, 
and more or less to affect 1,000,000 particles in a right ling in 
the direction of propagation, during the period occupied by a 
single particle in performing its individual motion in this par- 
ticular wave. On this assumption, the wave corresponding to 
the fixed line H must similarly embrace, in the direction of its 
length, 570,655 such particles. Suppose now a portion of the 
wether to be, from any cause, so compressed as to halve the inter- 
vals between its particles. When the waves B and H enter this 
compressed portion, B ought still to embrace 1,000,000, and H 
570,655 particles in their respective lengths; but the intervals 
