Two changes improved the apparatus over the one used at Potsdam [in 1886]. Morley suggested floating the heavy sand stone slab bearing the optical parts on mer cury. The stone, about 5 feet square, was mounted on a doughnut-shaped wooden block, which floated in a cast-iron trough containing the mercury. This method was economical and efficient, for the mercury-bearing removed practically all stresses and allowed the interferometer to glide smoothly around all points of the compass. Vibrations from outside disturbances, so troublesome in Berlin and even in Potsdam, were virtually eliminated.
With the new "interferential refractometer" Michelson and Morley were able to determine effects of the second order with an accuracy hitherto unobtainable. The second improvement lay in placing the mirrors so that the light was reflected over a path about 10 times longer than that of the earlier experiment.
While the stone, floating on its channel of mercury, turned steadily on its axis, one of them walked around it in a circle, keeping his eye on the moving eyepiece. The ob server could not touch the instrument, nor could he for a moment lose sight of the interference fringes. It was tiring to watch these through the small aperture of the eyepiece, only about a quarter of an inch in diameter, and the physical, mental and nervous strain told particularly on Michelson. As Morley said later, " . . . patience is a possession without which no one is likely to begin observations of this kind." The final readings were taken at noon, on July 8, 9 and 11 and near 6 p.m. on July 8, 9 and 12. They revealed no difference in the paths of light that could be interpreted as having been affected by the movement of the Earth through the ether.
Michelson concluded, exactly as he had after the 1881 ether-drift test, that the ether surrounding the Earth could not be at rest, as Fresneihad said it was, but must be whirling about, moving with the earth, so that it was not possible to measure the speed of the Earth against it.
The experiment had been eagerly watched, and as its impact struck the world of science, physicists reacted with astonishment and incredulity. However, reluctant as they were to accept the result, none of them thought of questioning either the experiment or the reliability of the verdict.
In the interim between the 1881 and 1887 ether-drift experiments, H. A. Lorentz had made a searching analysis of the whole problem. He explained the theory by comparing the space and time coordinates of a glass moving uniformly with the Earth's motion with those of a glass at rest. Space and time must readjust, much as they do everywhere on Earth where men go by local time, de rived from the time the Sun reaches a given location, rather than by absolute stellar time as it is calculated at Greenwich, England. Lorentz pursued his transformations into the electromagnetic field and found that his equations showed complete invariance between events in the moving glass and the glass at rest. This theory seemed to eliminate completely any idea of an ether dragged into motion with the earth, a theory that Michelson clung to because it explained the negative result of his experiment. Lorentz's ether stood even more firmly at rest than had Fresnel's. However, when Michelson's experiment was repeated in 1887 with more accuracy and the same negative result, Lorentz's stationary-ether theory was shaken to an ominous degree. But he, in turn, could not accept the idea that the earth dragged the ether along.
About 1892 George F. FitzGerald of Dublin visited Oliver Lodge in Liverpool to dis cuss the negative result of the Michelson-Morley experiment as well as the first-order effect of an ether-drift experiment that Lodge had made. FitzGerald said, "Well, the only way out of it that I can see is that the equality of paths must be inaccurate; the block of stone must be distorted, put out of shape by its motion . . . the stone would have to shorten in the direction of motion and swell out in the other two directions."
Lodge reported the substance of this conversation in Philosophical Transactions of the Royal Society of London in 1893. A length of 8,000 miles (approximately the diameter of the Earth), he wrote, would have to be shorted only 3 inches in order to ac count for the zero result of the experiment.
Lorentz came upon the same idea independently and embellished it with fervor. His hypothesis stated that every body which is moving at velocity u with respect to the ether contracts along the axis of motion by the factor
[the square root of (1-b2) = the square root of (1-v2/c2)]
This means that the Earth is squeezed very slightly into an elliptical shape because of its motion in orbit; that a ruler is shortened if pointed in the direction of this motion; and consequently Michelson's interferometer underwent this same modification only on the leg that was pointed in the direction of the earth's motion, thus canceling any hope of measuring a positive effect.
For different reasons, Michelson and Lorentz both longed for a positive result for ether drift. Michelson's frustration came to some extent from the feeling that mathematicians were taking the question out of his hands into a realm beyond his comprehension whence they drew their own, somewhat preposterous conclusions. He wished the dogged little Dutchman would drop the matter and accept the fact that both of Michel son's ether-drift experiments had been miserable failures. The ether was out there, but at present his instruments were unable to detect its presence. Lorentz, however, was not in a mood to let go of such a lively issue. With bulldog determination he wrestled with it until his fertile imagination produced an astounding explanation, which others also endorsed—that "rigid matter" may not really be "rigid." By the turn of the century, physicists would be calling this hypothesis the "FitzGerald-Lorentz Contraction." The idea was so shocking that only mathematical physicists could read about it without recoiling from the implications. The contraction theory was one solution to the question, and by this ingenious explanation Lorentz had managed to keep his stationary ether intact.
Michelson's disappointment at the failure of the second ether-drift experiment was harder to bear than at that of his first attempt, which had left some hope of finding a positive result with a different method. One of Albert Einstein's biographers later explained this disappointment.
"The famous Michelson-Morley experiment . . . proved conclusively that there are no different velocities of light! They are the same in all directions and their value is c, the speed of light, which strangely enough al ways remains true to itself, always constant, always unchangeable.
For the mechanist the result is catastrophic."