Talk:Cavendish experiment

Requested move: Torsion bar experiment → Cavendish experiment

This experiment is always referred to as the 'Cavendish experiment' and never the 'torsion bar experiment'; see the external links in the article, and:

• Encyclopedia Brittanica online article
• Harvard University Physics Dept. description
• S.F. State description
• U. of Saskatchewan undergrad lab experiment

I could only find one page where it was called something else: Michell-Cavendish Experiment. Also the current name is misleading; the apparatus used is called a 'torsion balance', whereas most people think of a 'torsion bar' as part of a car suspension. How about renaming this well-written article 'Cavendish experiment'? I've already added that name to the article. - Chetvorno 07:36, 1 April 2007 (UTC)

Done. —Nightstallion (?) 20:16, 12 March 2006 (UTC)

Fourmilab is bogus

The do-it-yourself experiment at Fourmilab ([1]) can't possibly be correct. A calculation using F=MA and F=GmM/(r^2) shows that it would take hours for the torsion balance to rotate as far as the video shows. Furthermore, a version of the experiment performed by Professor Norman Scheinberg (The City College of New York) uses much more massive weights (about 30 kg) and a much thinner wire (25 microns) made of tungsten - and the torsion balance only moved about 2.5 cm in twelve minutes. In the Fourmilab experiment, the movement due to gravity produced by the smaller weights and thicker, stiffer wire would be microscopic. (An earlier version of the article included a remark questioning Fourmilab's results, but it was removed.) If the video is not a hoax, the explanation for the rapid movement is almost certainly air currents created by the movement of his hands near the balance. The experiment by Professor Scheinberg is inside a box to prevent air currents from moving the balance, and Cavendish's original experiment was also protected from wind. - CronoDAS 03:46, 19 August 2006 (UTC)

Correction: Upon closer examination of the video, the torsion bar is actually touching the platform that one of the weights is placed upon. When he moves the weight, the foam shakes, knocking the styrofoam bar away. Furthermore, styrofoam is notorious for accumulating a large static electricity charge. All John Walker found was static electricity, not gravity. - CronoDAS 04:39, 20 August 2006 (UTC)

Although I haven't gone through the math, your criticisms of the 'Fourmilab' experiment sound valid. So glad someone has the experience to catch errors like that. I think I included the Fourmilab page as reference just because I was impressed that someone would try the Cavendish experiment in their basement with fishing line. But I should have noticed he didn't calibrate the apparatus by calculating the period of oscillation. I agree with removing the Fourmilab reference and replacing it with the Scheinberg experiment, a much better example of a homebrew Cavendish experiment. - Chetvorno 08:20, 1 April 2007 (UTC)

Add a picture

Could an artist please add a picture to illustrate this experiment as it would be so much easier to understand then.82.25.173.16 18:39, 3 December 2006 (UTC)

I added the original drawing from the experiment and removed the picture with dumbell type of torsion pendulum. That was confusing giving the impression that Cavendish used dumbells attached to the arm. In fact he had lead weights of 2 inches diameter suspended from each end of the arm. I will also add a detail showing the arm. Zeyn1 00:17, 20 August 2007 (UTC)

Great addition. --Chetvorno 19:29, 22 August 2007 (UTC)

Image of Cavendish

What kind of image is the one of Cavendish performing the experiment, and where did it come from? It looks for all the world like a manipulated photograph, but as far as I am aware photography didn't get started until the 1820s, 10 years after his death. Is it just an extremely good rendering? The image file itself doesn't seem to have any source information associated with it.

I believe that image is an illustration and it is misleading. I added "artist's rendition" to the caption. Zeyn1 00:59, 2 July 2007 (UTC)

Did Cavendish measure G?

I believe that Cavendish never measured or computed G therefore this sentence needs to be adjusted:

From the twisting force in the wire and the known masses of the spheres, Cavendish was able to calculate the value of the gravitational constant.

The historical references for this come from the two sources actually cited on the page. Clotfelter writes:

In 1798 Henry Cavendish performed an experiment now always described in physics textbooks as a measurement of the universal constant G. Cavendish did not report his work as a measurement of a gravitational constant, however, and in fact that did not become the standard interpretation for over 100 years.

I believe that to ascribe to Cavendish something he did not do is not an "interpretation," it is historical inaccuracy that needs to be fixed. Cavendish's biographers Jungnickel and McCormmach write in their book Cavendish:

What modern accounts usually say Cavendish did, he did not do. The universal gravitational constant he did not derive.

What is the consensus on this issue? I am also writing to Encyclopedia Britannica asking them to correct their claim that "Cavendish adapted Michell's design to make the first reliable measurement of G in 1798."

Also, G was defined in 1894 by C.V. Boys. See: C.V. Boys, On the Newtonian Constant of Gravitation. [Abstract] -- Proceedings of the Royal Society of London, Vol. 56. (1894), pp. 131-132. But Cavendish died in 1810. So Cavendish couldn't have measured G.

Since I am new in Wikepedia I would appreciate guidance about how to correctly edit this page to reflect the historical facts. Thanks. Zeyn1 22:16, 2 July 2007 (UTC)

Jungnickel and McCormmach's book was not in references so I added it. Zeyn1 22:22, 2 July 2007 (UTC)

Great point. You seem to be right. Cavendish's original paper can be read online at Googlebooks - A.J. McKenzie (1900) Scientific Memoirs Vol.9: The Laws of Gravitation. The gravitational constant doesn't appear in it, and there's no indication that Cavendish saw any importance in his experiment beyond finding the mass of the Earth.

At the time, Newton’s theory of gravitation was only 100 years old, and units of measurement had not caught up with it. I think with the system of units then in use, there was no ‘G'. In McCormmach & Jungnickel, p.337, it says "...he didn’t distinguish between weight and mass, so no gravitational constant appears" in his calculations. If weight (force) and mass are measured in the same units (pounds) then the gravitational constant G becomes numerically equal to $R_{e}^{2}/M_{e}$, where $R_{e}$ is the radius of Earth and $M_{e}$ is the mass of Earth. This confusion of units ties the gravitational constant, which has nothing fundamentally to do with the Earth, to the mass of the Earth. So to scientists of the time, ‘weighing the Earth’ was the same as determining G.

However, when the correct units are used, G can be easily calculated from his results:

$G=gR_{e}^{2}/M_{e}=3g/4\pi R_{e}D_{e}$, where $g$ is the acceleration of gravity at the Earth’s surface and $D_{e}$ is the density of the Earth. Using the value he obtained for $D_{e}$, 5.48 times the density of water, gives a value of 6.70(10^-8) cm^3/g s^2 for G, which is within 1% of the current value. So although Cavendish himself didn't determine G, his experiment did.

Cavendish’s experiment itself showed the need for a separate unit of force, pounds-force. Presumably, when that was adopted, Cavendish’s peers calculated a value for G from his data, although I haven’t been able to find when that was. Anyway, the article should certainly be corrected to say he didn't determine the gravitational constant. --Chetvorno 21:13, 30 July 2007 (UTC)

Rewrote article and added a section explaining that Cavendish didn't calculate G. Does anyone know when a value for G was first calculated? --Chetvorno 09:50, 7 August 2007 (UTC)

I removed broken link from external links section: http://wise.fau.edu/~jordanrg/bios/Cavendish/Cavendish_bio.htm. The article already has a link to Cavendish's Wikipedia bio. Zeyn1 14:09, 5 July 2007 (UTC)

Changes

I rather heavily edited the first few paragraphs to try and bring them in line with what's in the references. Some of the dates need checking and, perhaps, citations. I'll link the Henry Cavendish page to this one more explicitly. A Suggestion: It would be good to work in the fact that Cavendish observed the torsion pendulum's oscillations, not just deflections. - Astrochemist 04:06, 5 September 2007 (UTC)

Good rewrite. In general, I like your changes to the 'Did Cavendish determine G' section, and the new title. I agree with your suggestion that Cavendish's measurement of the torsion balance oscillation period should be included. The only thing I have an issue with is the first sentence. The first sentence should state for casual readers the modern significance of Cavendish's experiment. I think it misrepresents it to say just, ...[it] was the first accurate determination of the Earth's density. A reader who stopped there might think it only had significance for geology.

The experiment's modern importance is clearly more because it was the first accurate G measurement, than because it was the first accurate Earth density measurement. It's important not to exaggerate Clotfelter's point. Cavendish didn't determine G, but his experiment did. He presented his result as density because other gravity researchers Bouguer, Maskelyne, & Hutton did. Deriving G from his density result is trivial, and Cavendish would have done it if G had been in use at the time. In modern units, the experiment is purely a G measurement, and only yields the Earth's density when the Earth's radius and acceleration of gravity are introduced. This is why scientists since the 1890s have simply said that Cavendish determined G (Boys 1893). His experiment has nothing to do with the Earth; it could be done on the moon, and produce the same value of G (Boys 1893). His remarkably accurate results stood for 97 years, into the 'G' era, and became the standard value of G into the 20th century (Brittanica 1911). In fact, one reason that the 'G' form of Newton's law appeared (in the 1870s) was that Newton's law was being used more, because accurate calculations could now be made with it, because of Cavendish's value. His method was used to establish a lot of results in gravitational physics that have nothing to do with the Earth, such as the Eötvös experiment proving the equivalence principle. Many modern papers on G measurement start by citing Cavendish (U. of Wash), and G measurement conferences are held on the anniversary of the Cavendish experiment (Speake 1999).

One reason his results were accurate is because he measured the force between laboratory-sized masses that could be accurately weighed, rather than geographical features such as mountains. This was his other important innovation. After Cavendish showed it was possible, gravity researchers abandoned geographical methods for lab-scale methods, and many others were invented, such as the Jolly balance and Wilsing's method. Another result of his lab-scale experiment was to convince scientists that Newton's law of gravitation was universal; it applied to small-scale distances and objects like lead balls as well as planets. (Pratt 1836p.235)(MacKenzie 1900, p.vi)(Airy 1848 p.2). The gravitational interaction at small scales is today again an important research topic, in connection with string theory (Adelberger 2006).

For these reasons, I think the first sentence should say something about the experiment's importance to G measurement, and that it was the first measurement of gravity between laboratory-scale masses. As C. V. Boys said of the Cavendish experiment in 1892:

'Owing to the universal character of the constant G, it seems to me to be descending from the sublime to the ridiculous to describe the object of this experiment as finding the mass of the Earth or the mean density of the Earth...' Boys 1892 in Nature

--Chetvorno 10:01, 7 September 2007 (UTC)

Changed the first sentence, but added an explanation that Cavendish only determined the density. See supporting citations. At your suggestion added a para on Cavendish's measurement of the oscillation period of the balance. --Chetvorno 11:54, 7 September 2007 (UTC)

My first thoughts are that (1) readers should not be left to think (I'm agreeing with you) that Cavendish's work stops with geology but that (2) density should be mentioned first since that's what he actually measured. (I suppose that a philosopher of science would say that Cavendish actually measured times, masses, and distances and used a theoretical construct to get a density.) Let me read the above a few times and then I'll try to leave a better response. I'm familiar with the authors and material you mentioned. -- Nice job on the oscillation paragraph! -- Astrochemist 18:37, 7 September 2007 (UTC)

The Google version of Cavendish's 1798 paper has smears on two pages, so I had to dig out an original copy. Wish we could link to something better than Google's. — The paper's title contains the word "density" and the text does talk of measuring a force, albeit expressed only relative to weights. This suggests that density and gravitational force ought to be in the first sentence, since they are what Cavendish was after and what are in his paper. The experiment's reinterpretation and extension by later workers can quickly follow in a second sentence, with a comment on significance. Something like this:

The Cavendish experiment, done in 1797 – 1798 by Henry Cavendish, was the first laboratory measurement of the gravitational force between masses, and the first to use that force for determining the Earth's density. The greater significance of the experiment, as shown later by others, is that the results can give accurate values for the gravitational constant and for the masses of the Earth, the Sun, and other celestial objects.

That sort of wording is faithful to the historical record and shows the reader the broader and, to me, greater importance of the work, as emphasized by those who came after Cavendish. - Astrochemist 03:44, 10 September 2007 (UTC)

Questions and comments

Here are some other comments and questions.

I hope you don't mind, Astrochemist, I interjected my replies into your comments below. I felt your comments were important enough to merit separate discussions. --Chetvorno 09:10, 18 September 2007 (UTC)

Should the top paragraph be reduced and the historical content be moved to a new opening section called History or Background?

Should Henry Cavendish's picture be put in the article? There's only one that is genuine and it's at the top of the Cavendish page. (The original in the British Museum is somewhat better than what's on Wikipedia. Long ago I held it in my hot little hands. Wish I had taken a picture.)

Sounds great! So glad you have the experience to know which is genuine. I sure wouldn't. --Chetvorno 09:10, 18 September 2007 (UTC)

The sentence "Cavendish's measurement was remarkably precise for its time" suggests that Cavendish's values were clustered close together (precise). From the rest of the paragraph it seems that the intent is to say that his apparatus was sensitive (detected small forces). I've changed it.

They were both precise (clustered close) and accurate (close to true value), not just because of apparatus sensitivity, but also C's painstaking experimental technique, and his corrections for minor perturbations (most of his paper is devoted to this). You're right, the para didn't make that clear. I think the para should be expanded to mention the other two reasons for accuracy. --Chetvorno 09:10, 18 September 2007 (UTC)

The addition of a density equation at the bottom is good because the modern flow of information gives G first, then M_Earth, and then density. However, I'm not sure I've ever seen an upper-case letter used for density. My experience is that "d" or a Greek rho is common. Can either be used here?

I agree. "d" or rho sounds good to me. --Chetvorno 09:10, 18 September 2007 (UTC)

Where the text said "physicists would derive his results", I changed the verb to "use" since if we could derive Cavendish's density results (and the pendulum period, etc.) then we wouldn't have to do the experiment and a certain group of people might be out of work!

I meant 'derive' in the sense 'derive the number that's the goal of the experiment (G or density) from the raw numerical results', not in the sense 'derive from first principles'. Maybe 'analyze' would be a better word. The word 'use' is inappropriate and confusing; 'deriving' or 'analyzing' the results is an essential part of the experiment, and not considered a 'use' (IMRAD). I think the reason the GA reviewer failed the 'Derivation' section (below) was because of that word; he said the section was "not relevant" because he didn't understand it was part of C's paper. --Chetvorno 09:10, 18 September 2007 (UTC)

Hope you don't mind me jumping in here. To clarify my comment below, you have made a nice distinction between the actual experiment as Cavendish performed it, and the use that was later made of it. As you have it, a form of G was "thrown out" by Cavendish's calculations, but was treated by him as a number of no particular consequence. Later, when G was recognised as an important physical constant, it was realised that this number had actually been reliably calculated by Cavendish some years earlier as a by-product of his density experiment (similar in some ways to the recognition of e in mathematics).

My objection to the 'Derivation' section is that it is basically all about what Cavendish didn't do. In an article entitled 'Cavendish experiment', I would expect to find C's treatment of his results (ie his density calculation) rather than a modern reinterpretation of them... not that I have any problem with noting modern usage and a calculation for G and M_E. This is a tricky one though, as Cavendish does have the dubious distinction of being known for what he didn't do ;)

I hope this helps! Regards, EyeSerene^TALK 11:09, 18 September 2007 (UTC)

The phrase "Michell's torsion balance became the dominant technique" needs editing since a torsion balance isn't a technique. -- Astrochemist 03:32, 10 September 2007 (UTC)

You're right. I was trying to get away from the term 'standard equipment'. In gravity measurement the torsion balance has evolved quite a bit since Michell's time; there are spinning torsion balances, multiarmed torsion balances, asymmetric torsion balances, spinning external weights, etc. (examples). They're all different, only the principle is the same. Maybe something like: "Michell's torsion balance technique became the basis for most gravity measurements up to the present"? --Chetvorno 09:10, 18 September 2007 (UTC)

GA review

I have taken on Cavendish experiment for review under the Good Article criteria, as nominated on the Good article candidates page by Chetvorno. You'll be pleased to hear that the article meets none of the quick-fail criteria, so I will shortly be conducting an in-depth review and will post the results below.

Where an article is not an outright pass, but requires relatively minor additional work to be brought up to GA standard, I will normally place it on hold - meaning that editors have around a week to address any issues raised. As a precaution to prevent failure by default should this occur, if editors are likely to be unavailable over the next ten days or so, feel free to leave a message on my talk page so we can arrange a more convenient time for review. Regards, EyeSerene^TALK 17:22, 10 September 2007 (UTC)

GA fail

I have now reviewed this article under the six Good article criteria, and have commented in detail on each criterion below:

1 Well written FAIL

1.1 Prose

The article generally explains its scientific concepts well, and gives a good outline of the actual experiment. However, in places the prose could do with copyediting for grammar, flow and clarity. I have given some examples below, but the list is not exhaustive:

• "The Cavendish experiment, done in 1797 – 1798 by Henry Cavendish..."

Suggestion: "The Cavendish experiment is the name given to a series of experiments performed by British scientist Henry Cavendish from 1797-1798..." (note date ranges are separated by an unspaced en dash)

• "However, these were derived by others from Cavendish's result, which was a value for the Earth's density."

Suggestion: "Cavendish's goal was to calculate the density of the Earth; his results were later used by others to derive values for G and the Earth's mass."

• "The gravitational constant doesn't appear anywhere in Cavendish's paper..."

Suggestion: change "doesn't" to "does not" (avoid contractions unless they appear inside a direct quotation)

• "A further complication is that up through the mid-nineteenth century..."

Suggestion: "up through..." could be expressed less idiomatically; for example "until" (the article is about a British subject, so should really avoid exclusively US phraseology and spelling)

• "Equating the two formulas for torque..."

Suggestion: the plural of "formula" is "formulae"

1.2 Manual of Style

Whilst the article is well wikilinked and encyclopedic, it has a number of compliance issues with the MoS:

• Lead: this should summarise (not introduce) the article, and be able to stand as a mini-article in its own right (per WP:LEAD).

• Headings: these should follow the guidelines on WP:HEAD. Specifically, "The experiment" should be just "Experiment", and headings should not pose questions ("Did Cavendish determine G?").

• Citations: although not really a GA requirement, it is recommended that the templates on WP:CITET are used for references. They provide a consistent means of quoting refs, and can be processed by bots for things like tracking down dead webpages or converting ISBNs.

• Style: I think quoting measurements in their original imperial units for the actual apparatus is correct in this case, but they should also have an SI equivalent (see WP:UNITS for more details, but also note the optional exception to this general rule for scientific articles).

2 Factual accuracy PASS

The article is well-referenced, and makes good use of its sources.

3 Coverage FAIL

The subject of the article is comprehensively covered. A couple of points here though:

• "Derivation of G and the Earth's mass": Although I can see why this section has been included, it is not directly relevant to the article and detracts from its focus. As the title states, the subject of the article is the Cavendish experiment itself, so the article coverage should stick to this.

• Repetition: there are a few repetitive mentions of the fact that Cavendish did not determine G, this really only needs to be mentioned in two places (the lead and the section that deals with it).

4 Neutrality PASS

The article is neutral in its tone and contains no evidence of bias.

5 Stability PASS

The article history shows no evidence of recent edit-warring or major changes.

6 Images PASS

All images used are appropriately captioned (although "Artist's impression..." might be better than "Artist's conception" for the first image) and bear suitable copyright licenses.

Outcome

Although it is a promising and well-researched article, because there are quite a few issues to address - some of them fairly major - I do not believe Cavendish experiment can attain GA status in the short term; I have consequently failed it as a GA candidate. Once the comments above have been dealt with, please feel free to renominate the article on the candidates page for a further GA review.

If you have any concerns about the conduct of this review, you can list the article on the Good article review page for discussion by other GA reviewers. You can also leave questions or feedback on my talk page.

All the best, EyeSerene^TALK 12:16, 12 September 2007 (UTC)

No physical connection in Newton's idea

In Newton idea, there is no physical connection between point masses while in Cavendish experiment all masses (balls, spheres, wires, wooden rod, frame of torsion balance etc) are physically connected to each other one way or the other. Thus as a corollary the whole torsion balance is a raggle-taggle mass (balls, spheres, wires, wooden rod, frame etc). Therefore due to ramification of physical linkages, Sir Cavendish botched to represent the nitty-gritty authenticity of Newton’s gist.

By my reckoning, the gamut of the test can only be valid if done in space where there is no other local gravitational attraction including observer. 96.52.178.55 (talk) 20:56, 2 July 2009 (UTC) Khattak #1