In a spare moment, I was looking up the famous controversy involving Nature, Jacques Benveniste, and basophil degranulation. This was an experiment, reported in Nature, that purported to show that there could be a 'water memory' effect that could provide a mechanism for homeopathy. Essentially, the workers found that a substance diluted such that there should have been no molecules of it remaining still had a biological effect. There was naturally a lot of skepticism surrounding this paper, and Nature only published it on condition that they could send a team to Benveniste's lab to verify the results. The team found a number of statistical and methodological flaws in the study, and concluded that the results were erroneous. A useful summary of the controversy can be found here. Since then, a number of groups have failed to replicate the results (for example this one).
I came across a very interesting site belonging to George Vithoulkas (8 canards on the quackometer). It seems that not all homeopaths greeted the Benveniste study with unalloyed joy. Vithoulkas pointed out that the results of the study were the opposite to what would be expected from homeopathic theory. Apart from the idea that the 'potency' of a remedy is increased as it is diluted, the other unproven foundation of homeopathy is the 'law of similars'. This states that 'like cures like'; a substance that causes symptoms when taken in large amounts will cure the same symptoms when taken in homeopathic concentrations. According to this, if antibodies cause basophil degranulation in normal concentrations, the same antibodies should prevent it in homeopathic concentrations, the opposite to what Benveniste's team claimed.
Vithoulkas concludes that the 'memory of water' argument is a red herring, and has done nothing but damage the credibility of homeopathy. Given the recent fiasco surrounding the Homeopathy special issue on water memory, he clearly has a point. Even so, Vithoulkas isn't really engaging with the evidence. He just knows that the Benveniste research must have been wrong because it violated one of the (unproven) principles of homeopathy. What I find fascinating and brilliant about this is that plenty of homeopaths are happy to defend Benveniste's work, no matter the flaws, because it seems to support one of the basic tenets of homeopathy. For example, here's Dana Ullman on the study. In doing so, they miss that it completely undermines one of the other tenets.