5. Comments and Replies on the Memory of Water

Prof. R. D. Verma ( wrote on Mon, 30 November 1998
I am very much impressed by views on Benveniste controversy. Do you believe his results? How do you explain using your views of quantum mechanics? I know homoeopathy works. But how?

I :
Whether Benveniste’s results are reliable is, in my opinion, for biologists to decide. I do not know. I do not know either whether homeopathy works, although I hope it does, since the scientific implications would be fascinating. What I believe is that the theoretical objections that have been raised against Benveniste’s results are flawed. The argument that no molecule is left in the solution is, in my opinion, meaningless. The fact that the objections are flawed however is no proof that the results are correct. I have some conjectures on a possible theoretical explanation of Benveniste’s results.
As far as I know the best-known theoretical framework for the phenomena described by Benveniste is that of Prof. Giuliano Preparata and coworkers. There is book by him ( "QED Coherence in Matter" , World Scientific, 1995) with a chapter (Chapter 10, "Dynamics and Thermodynamics of Water") devoted to a theory of macroscopic "coherence domains" in water. Prof. Preparata concludes the chapter writing that "once we appreciate the considerable amount of coherence that characterizes our (i.e. Preparata’s, m.c.) theory of liquid water, it is not impossible to imagine that such marvellous ordered structure may retain and release electromagnetic information that it has acquired in some way or another". One may also read Prof. Brian Josephson’s letter to the New Scientist about these issues. I am just uneasy with the idea of a "marvellously ordered structure" storing, preserving and releasing electromagnetic information. But it's not impossible to imagine.  

John Benneth ( wrote on Thu, 18 February 1999
Really appreciate your article. What about nuclear magnetic resonance and liquid scintillation measurements of homoeopathic solutions, as is reportedly being done by Rolland Conte and associates at Medicine Quantale?
John Benneth

I :
Glad you liked it. Thanks for the links, which are indeed interesting, although the proposed methodology is still somewhat obscure to me. The possibility to use NMR to study highly diluted solutions is intriguing.

Thu, 13 May 1999. I : What follows is an e-conversation I had with a friend who works as a statistician at a pharmaceutical company. The friend prefers to stay anonymous, but she authorized posting her consideration. The tone is very casual since this was meant to be a private exchange, but the content is, in my opinion, quite interesting.

Friend :

I've read MOW stuff. I like none of these papers. Benveniste, for his part, is not very sound on experimental conditions (discarding some of the data subjectively). His critics [Maddox&al] are hysterical and ignorant. They raise a good point of subjectivity in recording the results, but even there make wrong statements. And the atmosphere of the visit and around it is disgusting. As to the latest paper [Hirst&al], you are right, it is strange. The authors have done a good experiment, clean from my statistical point of view. But they are bubbling around with the analysis. Unfortunately, their data is in color, and I can't see it on a photocopy [color pictures were then provided]. So, it is difficult to judge for myself. But one thing that I can not understand, may be I am missing the whole thing: at low dilution buffer solution works as well as succussed anti-IgE (Fig 3c). What does this mean? As to the analysis, they perform so many different tests, that adjusting for multiplicity, they should not have any power to detect anything except really gross differences. They are not clear on the hypothesis of their tests and on ANOVA models that they used. I would discard completely any of their analysis that do not take controls into account (ANOVA in Table1, probably). And their significant t-test seems to be Bonferoni adjusted. It is a very conservative multiplicity-of-test adjustment, and claiming "by chance" after that is very thin (though of course everything may be by chance, strictly speaking, even very strong results). Doing separate F-test for each treatment-dilution-session combination is GROSSLY bad, again because of power considerations (Table 2). Their "combined 'Fisher' P value" may be a right thing depending of what how they obtained that (They do not tell). Very strange (and does not add respect to the paper) that they do not give any numbers, only t- and F- statistics and P values. What about values for differences in counts (or percentages) for all treatments - dilutions - sessions, or summaries of these values? Plots are, of course, helpful, but with many points smooth curve through the data (or at least mean reference line) is needed to compare results with zero visually. Anyway, after reading I am not convinced of anything, moreover, I am convinced that they are not convinced themselves.

I :

I have a very simple question. The p=0.0027 in Table 2 for treatment A refers, as far as I understand, to the probablity that there is no actual difference between placebo and succussed high dilutions. This means that according to the data there is less than a 0.3% chance that placebo and succussed high dilutions are the same. Isn't this a very strong result? Doesn't this confirm Benveniste's result?


Yes, strictly speaking p=0.0027 is the probability of getting by chance the data they've got, if there is no difference between treatment A and control. Which is much more than enough to conclude that there is a difference. BUT: they do not tell what 'combined Fisher p-value' means and how they got it. The result depends not only on the null hypothesis, but also on the alternative hypothesis, and on the assumptions made. And there are so many inconsistencies in the paper... As to the pictures, they are striking. I think they have done a lousy job of conveying the results. Even, if one forgets about low dilution problem with placebo. High dilutions of three treatments (Fig 2), yes they are consistently different, and actually strangely monotonic, though one would expect them to be constant in both cases (with or without difference between the treatments). Now look at infinite dilution - i.e. control, the rightmost points. This is what they have to compare with for each session. They have exactly the same pattern. They are approximately equal to the mean of high dilution points with the same color (Treatment) and shape (Range). Actually except the highest range for treatment A, where the mean of controls is a little higher than any of individual dilution means. As to Fig 4, I can't understand what they are trying to say. As I understood different treatments always were done in separate session, so what 'single experiment' they show? Or color means something different here? Probably, the highest range of Treatment A is really different from others (from Fig 2 and from their talk about Fig3), and I think this what they detected by p=0.0027. But may be not the other two high dilution ranges. Generally, variability in controls is so large in comparison with the effect (see range of controls in Fig 2 versus difference between red square control and other red squares), that they should be very careful with the analysis to be able to detect them. And it's not surprising that analysis with different assumptions about variability give them different results.

On January 22, 2000 A.M.Selvam ( wrote
Chaos theory as applied to fluid flows can explain the observed "memory of water" . Fluid flows exhibit selfsimilar fluctuations on all scales of space and time. Selfsimilar spacetime fluctuations imply long-range correlations, that is, persistence or memory . The above website gives details of the theory in several published papers.

On January 23, 2001 Anthony Campbell( ; ) wrote
Thanks; an interesting discussion. I'll link to it on my site.
You may like to have a look at my electronic book "Homeomythology" (site as above)
Anthony Campbell

I :
Thanks to you.
Anthony's balanced and informative site is very interesting. Highly recommended.

On February, 2001 Ing. Drago Plecko, M.Sci. ( ; ) wrote
Dear Sir,
We are trying to produce the water that could "mimic" replacement drugs used in the treatment of heroin addicts. For that purpose, we tried to "imrpint" the well water with some "words" that could do the job, but haven't tested that yet. I would appreciate you to explain us how is the "sound" of a molecule recorded and later replayed to the water and could that be done with "heptanon" or "methadon".
Thank you very much.

I :
Thank you for your message. Unfortunately I cannot help you. I do not know how to "imprint" water with electromagnetic signals and I am not sure that it is possible to do so. I suggest that you address your questions directly to Dr. Benveniste at .

On March 29, 2001 Sam Nico ( ; ) wrote
I do not know whether this will interest you, but I have been working on some ideas that are based on a principle of the memory of nothing. In my thoughts, I had concluded that homoeopathy could only work if it contained nothing but memory, and that it would be made ineffective were a trace of actual material present. The ideas affect how we currently think of gravity, and there are important implications also for evolution thinking. It is by its nature mostly philosophical, but I believe that even a philosophical solution to the problem of quantum gravity is more important than one that can represent it in terms of equations. It also makes it more accessible, even though I feel it creates more problems than it solves. I would be happy to let you have a paper I am writing on this subject if you are interested.

HUmberto Avesani ( ) wrote on Saturday, May 18, 2002
I am an MD practitioner of Homeopathy in Argentina. I think that theories as such could make a revolition in the field of new ideas in Pharmacology and in in the science as a whole. Dr. Humberto Avesani

Submitted by Aaron () on Wednesday, October 13, 2004 at 19:58:40 :
comments: I don't know much of anything about quantum physics, but I know that the more particles an object consists of, the harder it is to keep it in superposition -- and antibody molecules are HUGE. Could they really stay in superposition long enough for quantum mechanics to have a noticeable effect?

I :
Thank you for your input, which brings into focus the relevance of the still hotly debated "quantum measurement problem" for the issues discussed here. I believe that your statement "the more particles an object consists of, the harder it is to keep it in superposition" is incorrect. Superpositions subsist as long as the system's evolution is unitary, i.e. until the sysyem's state is observed. This applies to any system, microscopic or not. Macroscopic superpositions are admittedly harder to detect, but that too can be and has been done (see e.g. [1],[2]), not only by basophils.


Submitted by mario ( on Wednesday, October 27, 2004 at 05:22:35

comments: I read this page and the comments on Dr Benveniste and his experiment. There's a lot of theoretical explanation going on. However, everybody seems to miss the main poit - there is nothing to explain. Benveniste repeated his experiment with scientists from Nature present. First he had succes, but then the people from Nature blinded the experiment (that is all they did) and it failed completely. Benveniste had been cheated by 2 of his assistants that altered the data to please him. Benveniste himself noted that only his assistant Jamal was able to get positive results, nobody else from his, including himself. Benveniste choose to ignore the possibility that Jamal was cheating or incompetent but chose to call it the 'Jamal-effect'. Getting quantum-mechanics and Schrodingers cat and all that involved is pointless, since there is no effect that needs explanation.

I :
You might have a look also at : and

Prof. Jacek Tyczkowski ( on Tuesday, April 5, 2005 at 09:20:37

Dear Dr. Vecchi,

I would like to give a reference in my paper to your text entitled "Red herrings in electrochemical high dilution" ( Has it been published in any journal? If so, please give me name of this journal, vol., pages and year.

Thank you in advance.

Best regards,

Jacek Tyczkowski

I :
Dear professor Tyczkowski,

thank you very much for your message. I apologise for my late reply. I am currently in the Indian Himalaya and internet cafes here are pretty thin on the ground.

I have not managed to publish my "electrochemical" remarks in a journal. Actually , since I got interested in fringe science, I have given up on getting published in "serious" journals.

I may suggest that you give a reference to my site, since I have set up things so that it may stay online for a while. Or you may give a reference to my unanswered message to Prof. Durst, which is archived by the Google newsgroup archive ([1]}, unchangeable and appropriately time-stamped. If you wish I will gladly post my page on a Google newsgroup so that it can be referenced.

Anyways, I truly appreciate your message and I will post it on my site, if that's ok for you, as I assume. I would also be grateful if you could provide me with a reference to your paper.

Best regards,

Italo Vecchi


Prof. Jacek Tyczkowski ( on Monday, April 18, 2005 wrote:

Dear Dr. Vecchi,

Thank you very much for your reply. In fact, I will give a reference to your internet site. However, I encourage you to publish your works in typical scientific journals. This way you will be able to provoke a broad discussion.

My paper entitled " In the search for physicochemical basis of homeopathy (unfortunately in Polish, only with an extended English abstract) will be published in Wiadomosci Chemiczne (journal of the Polish Chemical Society).

Happy climbing in Himalaya.

Jacek Tyczkowski

I :
Dear Prof. Tyczkowski,

I think that indeed giving a reference of my site is the best solution. In the next few days I will also post a copy of the current contents of my page on a Google newsgroup (*) so that in any case an interested reader may check the status of my page at the time of your reference.

I thank you for your advice and wish you all the best.

Italo Vecchi

(*) at

Anthony Campbell ( on Tuesday, May 31, 2005 at 04:30:52 wrote:

Could I ask you kindly to update the URL for my on-line book on homeopathy, "Homeopathy in Perspective". Many thanks.