Past observations by the astronomers such as Lacaillé, John Herschel and even H.C. Russell have suggested that the difference in magnitude was about three to four magnitudes, with the average of all WDS(96) observations being 2.49. The “average” magnitudes of the two components in the last 300 years has alleged risen in brightness, with relatively small changes in the difference in magnitude. (See Figure 3-1)
Today, it is clearly only about 1.5. My own thirty-one observations of estimating the difference in magnitude, between the years 1981 and 1990, range anywhere between 1.1 and 1.8, producing an average of 1.4. Hipparcos (1996) measured this difference as 1.36, so this is about right. Why this difference is so large by these early and notable observers remain still a mystery. This problem with magnitudes is also briefly stated in Burnham”s Celestial Handbook Vol.1.
Another issue, not mentioned by Burnham, is the way magnitudes were reported through most of the 19th Century. Unlike the modern magnitude system first set of by the variable, early observers tended to use “personal” magnitude scale, which could vary a much as six or seven magnitudes among visual astronomers. These variations were analysed in 1856 by Norman Pogson - the founder of the logarithmic magnitude scale that we use today. Here he found that the John Herschel (and to some extent William Smyth) underestimated the brightness of the first magnitude stars. (Hearnshaw., John, B.; “Origins of the Stellar Magnitude Stars”; S&T, 84, 5, p.494-499 Nov.(1992)) It is quite possible that the companion star has brightened in the last 100 to 200 years, though such changes would be, in many aspects, contrary to our current understanding of stellar evolution. The “A” components’ variability could also have an small influence on the estimates, but it should be average out instead of being consistently out by such a large amount. The actual colour of this star may have also changed from a deep yellow to a light-yellow / yellow, as reported by both Herschel and Russell. (3)
Minor variations in brightness for α Cen was reported in S&T, 76, 6, p.509 Dec. 1984 in the “News Notes”, first reported in the French professional Journal “Comptes Rendu”s in 7th June 1984 by the collaboration of several French astronomers. Lead by Eric Fossat, working in the European Southern Observatory in Chile discovered these small periodic variations over a six day observing window. Although the primary has the same spectral class as the Sun, being a G2 star, however, the star oscillates slightly in brightness every twenty minutes by c.0.65 magnitudes. Such oscillations are not uncommon in G-type stars, and such brightness fluctuations have been detected in the Sun, which varies in periods around five minutes. Since discovered in the Sun in 1965, the variations have been found in several stars using high-speed photometry centred on the narrow double-yellow Sodium lines in the combined spectrum. In the Sun, this is achieved using a bolometer - a very small telescope attached to an electronic photometer, and is very useful in measuring small changes in temperatures and energy fluxes. It seems that the “B” component is the culprit, whose data was subtracted from the combined light. Origins of these brightness fluctuations is unknown, but several projects in the field known as helioseismology, are slowly revealing the inner workings of the Sun and the how energy and material is transferred from the solar furnace in the core to the surface.