Critique and Response regarding Atomic Constants, Light, and Time
Question: Would please comment on this scientists review of your paper, “The Atomic Constants, Light and Time”.
"I have looked at the article carefully and come to exactly the opposite conclusion as the authors. The data they show is typical of a number that is actually constant. Early measurements have large errors and fluctuate either high or low (usually in only one direction) from later high precision measurements. Note in the article that from the early measurements the authors get a change of the speed of light of maybe 50 km/s per year. In the high precision measurements they get things like 0.005 km/s per year. This shows no real change but a completely understandable convergence of measurements. It happens all the time in science.
There was some real possibility that the speed of light may change about 3 years ago when supernova measurements showed a very small change in the fine structure constant. (This news made the New York Times). However, the size of the change was so small as to make the article you sent me totally false. In other words, recent measurements have measured the speed of light back billions of years and it is constant or changes only very slightly. However, even more recent measurements have made those of three years ago suspect. The very best experimental evidence is that the speed of light has been constant over the history of the universe, and that even if it has changed, it has changed by almost imperceptible amounts over the age of the universe. The data in the article is very typical of the history of a measurement of a constant value."
Setterfield: Thank you for the review. It covers old ground. Let me comment briefly about it.
In the first place a much more rigorous statistical analysis was done on all the data by Canadian statistician A. Montgomery and L. Dolphin and published in a peer-reviewed journal. That data analysis has withstood all criticism. The conclusion those authors came to was the same as in our 1987 Report, namely that the speed of light has dropped with time and that the associated constants were trending as our 1987 Report indicated. That paper by the Canadian statistician is “Is the Velocity of Light Constant in Time?”
Your critic here states that “The data they show is typical of a number that is actually constant. Early measurements have large errors and fluctuate either high or low (usually in only one direction) from later high precision measurements.” What I find interesting here is the insertion of the parenthesis statement (usually in only one direction). It is true that early measurements of a quantity may have large errors, but one thing that was noticed in the case where a constant quantity was involved was that the spread of data points was random around the fixed value. It rarely showed a one-sided departure. In the case of lightspeed, sixteen different methods were employed to measure c. In each case, without exception, the values obtained were always above the currently accepted value. This is not a normal distribution about a fixed point. If, as the critic claims, the data should show a one-sided departure, that would at the very most only pertain to one type of measurement method, and that would be different for the other methods. In short, a one-sided departure from a normal distribution is NOT considered to be the statistical norm for measurement errors.
The second point that needs to be addressed is that the measured drop in c was greater than the experimental error. There are 17 examples in the data set where the same equipment was used at a later date (sometimes by the same observer), and in each case the value of c was lower at the later date than for the earlier one. This is particularly important in the case of the data from Pulkova Observatory. There the data were collected by the aberration method for over 100 years. The data showed a decline in c. Interestingly, the Pulkova environment allowed a systematic error that shifted the value of c into a lower range than by other experiments. Nevertheless, the decline in c values was very apparent, and the early values for c were well above those currently accepted. In each case the decay was non-linear and tapering. At the same time another circumstance must be commented on. In the early 1880’s three different determination of c were made by distinctly different methods. The experimenters were unaware of each other’s activity. All 3 methods gave the same value of c to within 5 km/s, yet the value obtained was nearly 100 km/s greater than now.
This introduces the third point that needs to be made. Physicists of the 19th and 20th centuries admitted that the measured value of c was declining. In 1886, Newcomb commented in Nature for May 13th that the values of c obtained around 1740 were consistent among themselves, but placed c about 1% higher than in the 1880’s. In 1941 Birge made a similar comment about the data obtained in the 1880’s (some of it by Newcomb)) and commented that “These older results are entirely consistent among themselves, but their average is nearly 100 km/s greater than that given by the eight more recent results.” About that same time, Dorsey stated “As is well known to those acquainted with the several determinations of the velocity of light, the definitive values successively reported… have, in general, decreased monotonously…”
As far as the supernovae data are concerned, the quantity being measured was the fine-structure constant. That combines the quantities hc with the electronic charge and the permittivity of free space. It was shown in the paper under review by your critic that Planck’s constant, h, is inversely proportional to c, so that the quantity hc is invariant for all values of h and c. [Note that in 1965 J. H. Sanders commented in “The Fundamental Atomic Constants, p. 13, Oxford, that “Increasing values of h can only partly be accounted for by improvements in instrumental resolution and changes in accepted values of other constants.” A reviewer also commented that instrumental resolution “may in part explain the trend in the figures, but I admit that such an explanation does not appear to be quantitatively adequate.”] With the product hc being invariant, the variations in the value of the fine structure constant do NOT reflect a change in the speed of light, but rather a possible variation in the ratio of the electronic charge (squared) divided by the permittivity of free space in a gravitational field. Therefore, it is inappropriate to use this quantity as evidence for changes in the speed of light. The quantity hc has been measured as constant over astronomical distances by other methods. One cannot deduce any information about the behaviour of either h or c separately from such data.
I trust that gives some balance to the whole discussion.
Return to article
Return to Research
Return to GSR Home Page