next up previous contents
Next: Galilean Relativity Up: Problems Previous: Problems



Having postulated the existence of the ether as the medium in which light travels it becomes interesting to determine the properties of this material. First and foremost, since the light from distant stars does reach us, we must assume that the ether permeates the whole universe up to its farthest reaches. We must then imagine that the Earth plunges through this ether as it circles the Sun. The ether must then be very tenuous, for otherwise the friction would have stopped the Earth long ago. Let us now derive some other predictions derived from the ether hypothesis

As the Earth moves through this ether a kind of ``ether wind'' must be present on Earth's surface. To see why this should happen consider the following analogy. Imagine a windless day in which you take a ride in your red convertible which, unfortunately, has no windshield. As you speed up you will feel the air blowing, the faster you go, the stronger this wind is. In the same way, replacing air -> ether and red convertible -> earth , a very sensitive apparatus on the surface of the earth should detect and ether wind.

So, can the ether wind be detected? Apparently yes! The idea for the first experiments is based on the following argument. Imagine yourself back in your convertible (with no windshield) taking your nagging grandmother to the store; she sits in the back's safer. She talks all the time, but, fortunately, her words get blown back by the wind. In contrast she hears everything you say, for your words get blown back by the wind, right into her ears (good grief!). In the same way, as we stand on Earth, the ether wind should blow back the light coming from the stars. At different times of the year, the ether wind blows in different directions since the earth is moving in different directions, hence the observed positions of the stars should change (see Fig. 5.6)... and they do!

Figure 5.6: The shift in the observed position of the stars caused by the ether wind.  
\begin{figure} \centerline{ \vbox to 4.2 truein{\epsfysize=5.5 truein\epsfbox[0 -60 612 732]{5.eandm/}} }\end{figure}

But, wouldn't the earth drag with it some of the ether in its vicinity? Well, since this peculiar behavior of the images of the stars were observed, the earth must not drag the ether with it: ether goes through the earth ``much as the wind goes through a grove of trees'' (as described by T. Young.)

This consequence of the ether wind is not the only prediction of the ether hypothesis; in order to derive other consequences we need to go back briefly to Newtonian mechanics. Suppose you are in a train moving at a speed of 1m/s with respect to a train station. Suppose now you kick a ball in the direction of the train's velocity and which, as a result of your action moves at 2m/s as measured in the train. Then an observer in the station will see the ball move at 1 + 2 = 3m/s (see Fig. 5.7).

Figure 5.7: Addition of velocities according to Newtonian mechanics  
\begin{figure} \centerline{ \vbox to 3.5 truein{\epsfysize=6 truein\epsfbox[0 0 612 792]{5.eandm/}} }\end{figure}

Thus the two parallel velocities (the train's and the ball's with respect to the train) add up. In contrast if the ball were thrown up both observers would measure the same (vertical) velocity. Consider now the same situation but with light replacing the ball. If the train moves at speed v then light traveling forward will move at speed v+c. If the light-beam is directed upward both observers would measure the same vertical speed c. These conclusions are inescapable from the Newtonian standpoint and, because they are wrong, constitute some of the most important nails in the coffin of Newtonian mechanics. Let me examine first the following consequence derived from it.

Suppose you consider light going in air and that the same beam is made to enter a piece of glass. In air light will have a speed cair, while in the glass it will have speed cglass; these two quantities being measured at rest with respect to the ether. The experiment I want to discuss measures the ratio of speeds in glass and air. Now, if there is an ether, and the earth is moving at a speed v with respect to it, then one can select the orientation for the apparatus such that the beam happens to lie along the velocity v [*]. In this case the speed of light in air and in glass will be altered, they become cair+v and cglass+v respectively; the experiment should give the result (cair+v)/(cglass+v). If the beam is rotated 180o then the direction of the ether wind is reversed and the experiment ought to produce the value (cair-v)/(cglass-v). The amazing thing is that, as first shown by Arago, that this experiment gives the same value no matter how it is oriented with respect to the motion of the Earth through the ether. In order to explain this Fresnel suggested that transparent substances trapped some of the ether and dragged it along, and the amount and manner of trapping was ``just-so'' that the above experiment does not exhibit any effect. Of course the shift in the position of the stars would then imply that the air does not trap ether at all.

Curiouser and curiouser: the speed of light in glass depends on the color of light, nonetheless the above experiment gives no effect for any color. Therefore the ether trapped in glass should undulate with light precisely so as to compensate for this difference in speeds (note that the ether trapped with the glass travels with it).

So the ether is a medium which goes through all objects, but some of it is trapped by transparent substances and whose elasticity depends on the color of light going through it. In order to test this Fizeau performed a very important experiment. He sent light through tubes with water flowing in different directions. The water was supposed to drag at least some ether, which would then alter the speed of light. The results were positive and in accordance with Fresnel's hypothesis. So we have a big contradiction: the observation of starlight requires the Earth and the Earth's air not to drag any ether. But the Fizeau experiment requires transparent media to drag a significant (and measurable) amount of ether.

The most famous of the experiments made to detect the motion through the ether was the Michelson-Morley (or M&M)experiment. This is a very clear experiment. The idea is to send to take a light beam, to split it in two and send the daughter beams in perpendicular directions, these are then reflected back and recombined. The distances traveled by the daughter beams will be different and so there will be a mismatch between the two light wave trains resulting in a pattern of light and dark fringes after they are recombined (see Fig. 5.8)

Figure 5.8: A diagram of the Michelson-Morely interferometer  
\begin{figure} \centerline{ \vbox to 3 truein{\epsfysize=4 truein\epsfbox[0 0 612 792]{5.eandm/}} }\end{figure}

Now suppose we rotate the table where the experiment is placed. The speeds of the two beams with respect to the ether will change, and so will the times taken for the beams to recombine. Because of this the mismatch between troughs and crests in the two wave trains also changes and a shift in the pattern of dark and bright lines should be seen...except that it wasn't! No detection of the motion through the ether could be measured.

It was then claimed that the only thing proved was that the ether in the basement where the experiment was done was dragged along with the air. But the experiment was repeated a large number of times, in particular it was done on a hilltop: no effects were ever obtained.

This last result was the death blow to the ether theory: M&M's experiment showed that the ether must be dragged along by the air, while stellar observations denied precisely that!

next up previous contents
Next: Galilean Relativity Up: Problems Previous: Problems

Jose Wudka