FP's experiments discussion

  • Ascoli, "The major evidence in this paper is that its authors have disavowed the previous MF declarations about the absence of foam in the F&P open cells."


    Well, NO, they have always been aware of foam and bubbles , as described by Fleischmann [ 1]

    "We once had a batch of D2O that foamed badly. We traced this back to the Girdler-Sulfide process used by AECL (Atomic Energy of Canada Ltd.)."

    [1] http://lenr-canr.org/acrobat/Fleischmanlettersfroa.pdf


    Your reminding of MF awareness of the foam problem has the only effect of worsening his responsibilities with respect to the 1992 results.


    Regarding the quote, it didn't come from MF, but from McKubre (MCHM):

    From: http://lenr-canr.org/acrobat/Fleischmanlettersfroa.pdf

    2005-02-24

    Bury Lodge heading

    Dear Mel and Linda,

    […]

    If you check back through the correspondence, you will see that I eventually decided that some of the bottles of D2O had been contaminated by HDO. I must admit here that I was influenced in reaching this conclusion by our early experiences with the start-up of the N.H.E. Project (1). There was clear evidence of the “quenching” of excess enthalpy generation in the key experiments due to the addition of HDO instead of D2O (actually it looked more like the addition of H2O to me!). You have also pointed out that the anomalous value of (kR′) on day 61 was probably due to foaming in the cell. This is another problem which we pointed out to our Japanese colleagues. Samples of D2O sometimes contain added detergent to aid the filling of the NMR tubes! 146

    […]

    146 MCHM The comment is more of memory aid for me. We once had a batch of D2O that foamed badly. We traced this back to the Girdler-Sulfide process used by AECL (Atomic Energy of Canada Ltd.). We did not confirm this but I suspect that a lot of heavy water – and a lot of heavy water experiments – were heavily affected (not positively) by the presence of detergents (for whatever reason).

  • Ascoli,


    Again: Fleischmann were very well avare of possible foaming.


    Page 657:


    "

    Now as to the possible foaming in the cells. The Japanese were plagued by this problem due to their use of D2O destined for N.M.R. experiments. This contained added detergent to aid the filling of sample tubes. We wrote to them at length about this and I thought that the problem had been cleared up.

    I am getting back into my old ways! Regards also to Linda. How is your teaching getting on? Best wishes

    Martin

    "


    https://www.lenr-canr.org/acrobat/Fleischmanlettersfroa.pdf

  • The paper of F&P together with their previous papers referred to in the paper provides enough information to replicate the study by other scientists.

    [...]
    One example in this case is then the precise replication of Lonchampt, which indicated the same range of results..


    Almost no one was able to replicate the F&P experiment reported in their 1992 paper. The only one who claimed to have partially replicated it was Lonchampt, whose attempt was closely followed by F&P, otherwise he probably couldn't have been able to repeat the experiment on the basis of the scanty information contained in the F&P paper. Moreover, Lonchampt used a different method to calculate the energy balance, which is quite indicative of the inadequacy of the F&P method reported in the paper we are talking about.


    Anyway, the alleged excess heat claimed in the Lonchampt paper presented in 1996 at the ICCF6 has almost disappeared in the subsequent years, which clearly indicates that the first positive results were due to miscalculations.


    Quote

    The purpose of Scientific papers is not to provide all raw data so that other scientist can recalculate all their results in that particular study.

    […]

    Ref. On purpose of papers;

    https://writing.colostate.edu/guides/guide.cfm?guideid=83


    Well, the section "Results" (1) of the Guide you have linked says "Detailed data--measurements, counts, percentages, patterns--usually appear in tables, figures, and graphs". The data provided by F&P are not at all detailed.


    (1) https://writing.colostate.edu/…fm?pageid=1563&guideid=83

  • Again: Fleischmann were very well avare of possible foaming.


    Page 657:
    "Now as to the possible foaming in the cells. The Japanese were plagued by this problem dueto their use of D2O destined for N.M.R. experiments. This contained added detergent to aid thefilling of sample tubes. We wrote to them at length about this and I thought that the problem hadbeen cleared up. ...


    OK, now the quote is really from MF. But I can't understand what you are trying to say.


    I fully agree that MF was very well aware of the foaming issue. He knew his cells produced a lot of foam. The video published by Truthloader (1) caught him while showing his foamy cells to a Japanese man, as also reported in a previous jpeg (2). So, the only possible conclusion is that he was also aware that the brightness inside the cells during the final boil-off period, which was presented as boiling water in the F&P paper to ICCF3, was actually due to the foam.


    (1) https://www.youtube.com/watch?v=3OQu44UIC_s

    (2) FP's experiments discussion

  • Almost no one was able to replicate the F&P experiment reported in their 1992 paper. The only one who claimed to have partially replicated it was Lonchampt, whose attempt was closely followed by F&P, otherwise he probably couldn't have been able to repeat the experiment on the basis of the scanty information contained in the F&P paper. Moreover, Lonchampt used a different method to calculate the energy balance, which is quite indicative of the inadequacy of the F&P method reported in the paper we are talking about.


    Anyway, the alleged excess heat claimed in the Lonchampt paper presented in 1996 at the ICCF6 has almost disappeared...


    Do you have any documentation to indicate how many tried to replicate the 1992/93 papers?


    I think most scientists where busy developing their own setups and researching the causes of cold fusion as such, not If excess heat increased at higher temperatures.


    The papers from 92 and 93 and their references where mor than enough to replicate the F&P setup. But of course, you are entitled to your opinion of "Scanty information " 😉


    Lonchampt, as I said several times, recognised the difficulty in identifying water levels, so choose to calculate from the last fill until dry cell.


    And confirmed the F&P main hypothesis, that energy and power density increase with temperature.


    It is not important what the excact figure is. it may be possible Fleischmann was off, but not as much as you think.

  • I think most scientists where busy developing their own setups and researching the causes of cold fusion as such


    In the 1992 paper F&P achievedspecific xsheat intensities of up to 21.4.with high current densities of up to 512.

    Mengoli et al achieved lower intensity of 4.33 with a current density of 121.


    The focus of Mengoli was'95C isothermal conditions without boiling and the effect of cathode geometry inter alia.

    However their results at lower current densities were consistent with F&Ps.


    It was obvious that intense boiling conditions,/high current densities were achievable(with difficulty) for only short periods

    so researchers were looking at other ways to increase the rate of reactions.


    The Xs heat had already been confirmed... the research was not involved with confirming Xs heat but

    making it easier/cheaper to achieve and more reliably.


    Journal of Electroanalytical Chemistry 444 (1998) 155-167

    Calorimetry close to the boiling temperature of the D20 jPd electrolytic system

    G. Mengoli ,, M. Bernardini , C. Manduchi , G. Zannoni

    https://www.google.com/url?q=h…Vaw1KoWgsy045d7nJ3No5m2-v

  • Apropos of Lonchampt


    In the ongoing debate on the interpretation of the F&P results of their 1992 boil-off experiment, the results obtained by Lonchampt and reported in his 1996 paper to ICCF6 (1) are often presented as the best confirmation of the correctness of the results claimed by F&P. In reality, they only confirm that the F&P method gave wrong results.


    On Table 2 of his 1996 paper (1), the results obtained with his method - which applies the energy balance to the entire boiling period - are listed in the penultimate column titled "Mean relative excess heat". For the tests with Pd samples, they range from 3 to 20 %, but these percentages are referred to the "Available Enthalpy", which is the part of the "Enthalpy input" remaining after subtracting the "Enthalpy losses". This is a not correct way to calculate the relative error (he called "Excess heat"), because in this way the single inaccuracies are added up. The "Mean relative excess heat" should instead have been calculated with respect to the total "Enthalpy input" and the resulting percentages would have ranged from 1.6 to 8.3 %. This level of energy imbalance is easily explainable by normal experimental errors, especially considering that Table 2 lists all and the sole 5 positive tests, as reported in the paper (1): "Since the beginning of our experiments, in 1993, 18 runs have been carried out. Only five of them have produced excess heat, with high purity palladium cathodes."


    Interesting to notice that, among the only 5 positive tests, one was carried out at the beginning of 1996, while the remaining 4 were carried out in September/October 1996, i.e. a few weeks or even days before the ICCF6 conference, which was held in Japan starting from October 13, 1996, in where these results should have been presented. What an extraordinary coincidence after 3 years of almost complete failures!


    Furthermore, the data reported in Table 2 concerning the first of these 5 tests contain an incredible error. In fact, the "Available Enthalpy" - which is the "Enthalpy Input" minus the "Enthalpy losses" - was set at 156876 J instead of 164400 J. This error propagated in the columns on the right, so that the reported "Excess Heat" is 24623 J instead of 17099 J, causing an erroneous increase of the relative energy imbalance from 10 to 16% (from 4.1 to 5.9 % if correctly referred to the "Enthalpy Input"). If the very scrupulous French engineer made such a blatant error on a test performed several months before the presentation of his paper, I wonder which reliability the results claimed for the other 4 tests carried out just before the beginning of the ICCF6 conference could have.


    Until now, we have examined the results obtained by the Lonchampt method of considering the entire boil-off period. Now it's time for the "Grand Finale".


    Let's talk Lochampt himself (1): "As described in section 3.2, the exact evaluation of the excess heat can be made only at the end of the experiment, since it is difficult to follow accurately the water level during the experiment. However it is very likely that most of the excess heat occurs at the end of the experiment after the voltage burst. We call this last period the “grand finale.”"


    This phrase means that after having obtained a low energy imbalance (which he called "excess heat") with his overall method, he changed the calculation method and restricted the energy balance to a much shorter period called the "grand finale", obtaining in this way the much higher "Relative Excess Heat during "grand finale"", which he reported in the last column of Table 2.


    Lonchampt doesn't describe what this mysterious "grand finale" is and which criteria he used for calculating the relative excess heat, apart from a generic "it is very likely that most of the excess heat occurs at the end of the experiment after the voltage burst". Very likely for whom? Anyway, the following review of the ICCF6 provides a useful hint for guessing the source of this second method:

    From http://www.infinite-energy.com/images/pdfs/6thiccf.pdf - Rothwell on Infinite Energy, October 1996


    … Lonchampt is a CEREM commissioner, and an engineer not a scientist, (thank goodness). These experiments are exact replications of the 1993 boil-off experiments reported by Pons and Fleischmann in Physics Letters A176. This is exactly what cold fusion cries out for: careful, step by step replications done by people who follow directions. Biberian said that he and the other scientists in the project wanted to incorporate various “creative improvements” but Lonchampt insisted on doing a precise replication with assistance from Pons and Fleischmann. That is why it worked, …


    "That is why it worked": because he did "a precise replication with assistance from Pons and Fleischmann". It probably means that the reported excess heat of about 150%, much closer to that claimed by F&P in their 1992 paper, derives from the application of the same fake method of calculating the excess heat only in the final few minutes of boil-off stage, which was applied by F&P in their 1992 experiment.


    This fantastic result was also reported in the Lonchampt conclusions (1): "Our results concerning the relative excess heat (percentage of excess heat to enthalpy input) can be summarized as follows: […] - at boiling, up to 150% especially in the final phase which appears as the best condition to get a large amount of excess heat." This was a false statement, because the true excess heat that Lonchampt got with the method he explained in his paper was the 20 % (which actually was 8.3 % when referred to the total "Enthalpy Input") reported in the penultimate column of the penultimate row of Table 2.


    His unscientific choice to report results obtained with a method (the "grand finale") not explained in the paper, caused the Lonchampt replication to be widely mentioned by the F&P supporters for the 153 % shown in the second row of the last column of Table 2 (1), not considering that:

    a – the data of the second row contain an incredible error of elementary calculus;

    b – the data of the last columns were probably calculated by using the F&P method, which Lonchampt himself had criticized (1): "It is difficult to follow accurately the level of water during this period because of the formation of foam, so it is only at the end of the experiment, when the cell is dry that the excess heat can be calculated with precision."


    In conclusion, the Lonchampt paper only demonstrates, besides of how sloppy and inaccurate are the more celebrated CF documents, how big the error can be by using the calorimetric method adopted by F&P in their 1992 paper.


    (1) http://www.lenr-canr.org/acrobat/LonchamptGreproducti.pdf

  • Do you have any documentation to indicate how many tried to replicate the 1992/93 papers?


    I only know that the general opinion in the LENR community is that Lonchampt has been the only one to have successfully replicated the 1992 boil-off experiment:

    From http://coldfusioncommunity.net…mns/v21/1_JCMNS-Vol21.pdf - Biberian, 2016


    Georges Lonchampt was one of the few French researchers who, from day one, worked on Cold Fusion. He started performing his own experiments, and later worked with Martin Fleischmann and Stanley Pons. He successfully reproduced the two scientists’ original experiment, and was the only one able to successfully replicate the boil-off experiment.


    From https://theierecosmique.com/20…e-ou-arnaque/#comment-613 – AlainCo, September 8, 2016


    Un papier à relire sur F&P est celui de George Lonchampt, un maniaque reconnu, le seul a avoir répliqué non pas seulement le phénomène, mais la calorimétrie, et donc compris sa sensibilité et sa subtilité.

    http://www.lenr-canr.org/acrobat/LonchamptGreproducti.pdf



    Quote

    he papers from 92 and 93 and their references where mor than enough to replicate the F&P setup. But of course, you are entitled to your opinion of "Scanty information "


    Oh, sure. These papers were so accurate to allow a replication of the F&P setup, that even the cell section on Figure 1 was wrong!

    See: FP's experiments discussion


    Quote

    Lonchampt, as I said several times, recognised the difficulty in identifying water levels, so choose to calculate from the last fill until dry cell.

    And confirmed the F&P main hypothesis, that energy and power density increase with temperature.


    Please, look at the previous comment: FP's experiments discussion

  • Apropos of Lonchampt (part 2)


    As mentioned before (*), an extraordinary feature of the Lonchampt results presented at ICCF6 (1) was that 4 tests, out of the only 5 positive ones, were carried out in the last weeks before the conference. Such a coincidence is suspiciously unlikely.


    Leaving aside the need to provide positive results for participating to an attractive event, one possible explanation of this lucky sequence of positive tests lies in the choice of the parameter KR. In fact, as shown in Figure 5, the KR chosen for the September 30 test, which gave the maximum "excess heat" (an alias for "energy imbalance"), was 0.0905 (a wrong notation, which has to be understood as 0.0905x10-8 W/K4), a value greater than those reported in Figures 3 (0.078) and 4 (0.075). For equation (2), the heat losses by radiation - when the cell is at 100°C and the bath is at 20°C - amount to 10.66 W. This datum allows to calculate the duration of the boiling regime, i.e. the period during which the cell temperature was greater than 99°C, starting from the "Enthalpy losses" listed in Table 2. Dividing 178700 J by 10.66 W, the duration of the boiling regime results 16762 s, or 4h40m. Assuming that a better choice for the KR would have been 0.078(x10-8) – the value estimated during the calibration experiment held on June 10, 1996 and reported in Figure 3 - the heat losses would have been 9.19 W and the "Enthalpy losses" during the boiling period would have amount to about 154000 J. It follows that the "Enthalpy used for vaporization" would have been much greater than the 129900 J reported on Table 2, namely 154600 J. This last value almost equalize the 155570 J required to vaporize all the 72 cm3 of electrolyte present in the cell at the onset of the boiling regime, so that the energy imbalance would have been only 970 J, that is the 0.3% of the "Enthalpy input".


    In conclusion, of the 5 positive tests listed on Table 2 - out of the 18 tests with Pd samples carried out during a research activity lasted more than 3 years - the first is affected by a large calculation error, as explained in (*), and the remaining 4, in addition to have been carried out in hurry on the last moment, owe their apparently positive results to a KR value much higher than the value derived in the sole calibration test documented in the Lonchampt paper.


    (*) FP's experiments discussion

    (1) http://www.lenr-canr.org/acrobat/LonchamptGreproducti.pdf

  • Well, Ascoli,


    You have many comments here, but now it's Christmas and I have other Priorities 🤓. I will look at your comments later.


    But just a few general remarks,


    The reason I have referred to Lonchampt, is that I am not awar of other attempts to replicate.


    I believe we would have heard oh Those attempts lf there where any, but as I have said numerous times, the 92/93 papers are not the most important ones.


    Most CF scientists have been busy testing out their own ideas to trigger LENR, not copy F&P. The most successful in D/Pd systems are the co-deposition research at SPAWAR. They got very high sucessrate, so I do believe there are other LENR setups that are better for CF research than the F&P electrolysis line.


    The fact that Lonchampt only got excess heat from some 5 of 18 cells, shows the difficulty in acheieving the right conditions in F&P type setup.



    And as I said earlier it's not the numbers themselves that matter, but the trend, does Excess heat increase at higher temepratures or not.


    Testing different types of hypothesis is the road to achieve higher success rates.


    And now you also need some vacation Ascoli 😉.


    Have a merry Christmas to you and all others on this forum !🦌🦌🦌🦌🌲🌲🌲

  • You have many comments here, but now it's Christmas and I have other Priorities 🤓. I will look at your comments later.


    But just a few general remarks,

    [...]
    And now you also need some vacation Ascoli 😉.


    Have a merry Christmas to you and all others on this forum !🦌🦌🦌🦌🌲🌲🌲


    Thank you, Lande. :)

    Merry Christmas to you too and to all the L-F members and guests.


    I will wait your next post to answer your last remarks.

  • Tell the site ICCF 22, otherwise I can not find the information.

    Нефть - это кровь планеты, надо сделать модель планеты и мы получим генератор Тарасенко, эта энергия покорит вселенную! :lenr:

  • https://lenr-canr.org/index/Pr…atter%20Nuclear%20Science


    This is for reference for when the battle begins again. Rothwell did a nice job putting all the ICCF's under one roof (on one page), with links to most presentations for easy research. Lots of good info in there. In just a few minutes, I already see Amoco replicated FP's right after the announcement. Ascoli is going to have his hands full discrediting each and every one.

  • I wish an Happy New Year to all the coldfusionners all over the world.

    Best Regards


    Fabrice.

    Happy New Year to you too! We must find a new energy, the planet is dying!

    Нефть - это кровь планеты, надо сделать модель планеты и мы получим генератор Тарасенко, эта энергия покорит вселенную! :lenr:

  • Proposed poll: Boiling Water or Foam?


    Year we are: 30th anniversary has come. Nice year to everybody! :)


    After the last 2018 polls, I would propose to the L_F administrators to dedicate the first 2019 poll to F&P and in particular to the controversy concerning the interpretation of the videos which document their 1992 boil-off experiment.


    To simplify the poll questions, let's make reference only to the video published by Krivit on YouTube in 2009

    https://www.youtube.com/watch?v=mBAIIZU6Oj8


    This video shows the time-lapse recording of the 4 final phases of the 4-cell boil-off experiment carried out in April-May 1992 and whose results and claims was reported in the F&P paper presented at ICCF3, held in Nagoya in October 1992. As explained in this paper, video recording is an integral part of the experimental data gathered during this experiment and its analysis allowed the experimenters to estimate a time of 11 minutes for the complete boil-off of the last half of the water content:

    From http://lenr-canr.org/acrobat/Fleischmancalorimetra.pdf (bold added)


    ABSTRACT

    We present here one aspect of our recent research on the calorimetry of the Pd/D2O system which has been concerned with high rates of specific excess enthalpy generation (> 1kWcm-3) at temperatures close to (or at) the boiling point of the electrolyte solution. This has led to a particularly simple method of deriving the rate of excess enthalpy production based on measuring the times required to boil the cells to dryness, this process being followed by using time-lapse video recordings.


    [other paragraphs]

    […]

    For the second value of the pressure, 0.97 bars, the cell would have become half empty 11 minutes before dryness, as observed from the video recordings (see the next section) and this in turn requires a period of intense boiling during the last 11 minutes.

    […]

    The simplest procedure is to make time-lapse video recordings of the operation of the cells which can be synchronised with the temperature-time and cell potential-time data. Figs 6A-D give the records of the operation of four such cells which are illustrated by four stills taken from the video recordings, Fig 10A-D.

    […]

    As it is possible to repeatedly reverse and run forward the video recordings at any stage of operation, it also becomes possible to make reasonably accurate estimates of the cell contents.

    […]

    Figure 10. Stills of video recordings of the cells described in Fig 6 taken at increasing times.


    The blue arrows superimposed on the video follow the lowering of the brightest part of the cell contents which, in the F&P interpretation, represents the residual boiling water that is rapidly vaporizing due to the excess heat produced by the extraordinary F&P effect occurring in the submerged cathode. The amount of this claimed excess heat was calculated on page 16 of the F&P paper assuming that the last 2.5 moles of water (i.e. half of the initial 5 moles) vaporized in 600 s.


    The alternative interpretation is that the lowering brightness shown in the video, followed by the blue arrows, is essentially due to a thick layer of foam accumulated over a thin layer of residual water kept boiling by the normal Joule effect due to the electrolytic current still flowing in the cell. See: FP's experiments discussion


    Since its appearance on the debate about the 1992 F&P experiment (FP's experiments discussion), this alternative interpretation of the video aroused an intense discussion whose participants expressed many different positions. The proposed poll would invite the L_F members to express their opinion on this controversy by choosing one of the following 5+5 positions:


    A – No guess on the video because:

    A1 – The only valid interpretation is that provided at the time by F&P, who were well aware of the foaming problem;

    A2 – It is unfair to dispute an experimental work when its authors are no longer able to respond to criticisms;

    A3 – The video is too deteriorated to allow any meaningful evaluation of its images;

    A4 – The topic no longer matters, because, independently of the correct interpretation of the video, the FP effect has been experimentally confirmed by other authors;

    A5 – The argument is too much important to be judged by non-professionals and requires a in-depth investigation by expert equipped with suitable video-processing tools.


    B – The best interpretation of the video is:

    B1 – Certainly, or almost certainly, the brightness in the cells is mostly due to boiling water;

    B2 – The boiling water interpretation is more likely than that of the foam;

    B3 – Boiling water and foam have about the same probabilities;

    B4 – The foam interpretation is more likely than that of the boiling water;

    B5 – Certainly, or almost certainly, the brightness in the cells is mostly due to foam;

  • Another poll and we may have a mutiny on our hands! So no thanks for now. But if we ran one, I would have to select A1/A3.


    "B" is hard, because I saw one segment where it looked like the bubbles I see coming from the bottom of a glass pot, just before it begins a rapid boil. In another I saw foam, and yet another I saw what looked like condensate through frosted glass. Then, I could have sworn I saw Elvis at one point, but only briefly.


    And should you not have a "C1" -Does not matter what you saw, because entrainment was not a factor?

  • Another poll and we may have a mutiny on our hands! So no thanks for now.


    Take your time. There are almost 3 months before March 23.


    The poll on the "foam issue" would be different from the last two. It's more technical. The boil-off videos are the main experimental evidence left by F&P of their activity on cold fusion. It is considered self-convincing by a LENR expert such as Rothwell (1): "As far as I am concerned, this is first principle proof, and it is as convincing as a self sustaining machine, or as Fleischmann's boil-off video."


    At that time, a few days after the Ecat demo held on January 14, 2011, he was comparing the F&P boil-off video with those released by Rossi. The following year, Rothwell published the "IMRA time-lapse" video, accompanied by these words (2): "This original Fleischmann and Pons video shows the dramatic heat effect of the cold fusion reaction on the water fuel." This longer version of the F&P video was coupled with the other video, called "Brief Introduction to Cold Fusion", whose purpose was to "provoke a viral reaction attracting thousands of viewers. Possibly even millions" (3).


    It means that the boil-off video was considered very important and easy to interpret, even by many people not expert in the field. So, why not ask the opinion of the few experts which follow the most important forum dedicated to CF/LENR? The 30° anniversary is a good opportunity to do it.


    Quote

    But if we ran one, I would have to select A1/A3.


    Fine, but you should choose the most decisive for not giving any guess.


    Quote

    "B" is hard, because I saw one segment where it looked like the bubbles I see coming from the bottom of a glass pot, just before it begins a rapid boil. In another I saw foam, and yet another I saw what looked like condensate through frosted glass. Then, I could have sworn I saw Elvis at one point, but only briefly.


    If you chose one "A", you can't chose any "B". If you chose "B" and you think that the pro-boiling water evidences are almost equivalent to the pro-foam ones, then you can vote B3.


    Quote

    And should you not have a "C1" -Does not matter what you saw, because entrainment was not a factor?


    No, it's not a valid option. Entrainment has nothing to do with the "foam issue", as I have already explained to you (4).


    Your poll is one of those theoretically good ideas, like a 3-way STV Brexit people's vote (sorry, a UK political reference) that probably won't happen due to most people not liking complexity.


    The number of the options of the F&P's poll are similar to those of the last two polls and it's not necessary to have a majority or make a decision on the basis of its outcome.


    (1) https://www.mail-archive.com/v…@eskimo.com/msg41533.html

    (2) https://coldfusionnow.org/flei…roduction-to-cold-fusion/

    (3) https://www.mail-archive.com/v…eskimo.com/msg101767.html

    (4) FP's experiments discussion