Talk:International System of Units/Archives/06/2020

This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Note: This long thread soon morphs into whether the standard atmosphere is a metric unit, which branches into whether the torr is a metric unit. --Reuqr (talk) 19:05, 30 May 2020 (UTC)

I was not questioning the status of g as a unit, but as a metric unit. Inclusion in IEEE/ASTM SI 10-2010 does not justify the claim (if it were we would be claiming the gallon is a metric unit). I think it should be removed from the list. Dondervogel 2 (talk) 18:25, 18 May 2020 (UTC)

Of course the gram is a metric unit, and of course the proper SI symbol for it is "g". If you wish to discuss this further I suggest you provide a quotation from the article that shows what you are concerned about, because searching on a single letter of the alphabet is useless. Jc3s5h (talk) 19:58, 18 May 2020 (UTC)

The {{dubious}} tag was on "g (as a unit)": see here. —Quondum 20:06, 18 May 2020 (UTC)

One argument is that g accepts metric prefixes and freely mixes with other metric units such as the hertz. For example, in this paper, we see all of the following (and the context makes it clear it is the standard gravity and not grams): micro-g, nano-g, μg, mg, μg/√Hz. It is easy to find other reputable sources that use this terminlogy by googling ‘ "acceleration" "micro-g" ’, e.g. this one, this one, etc.

It is true that g could also be an inch-pound unit (really, it could be a unit in any system of units), but all that follows from this is that it is both an inch-pound unit as well as a metric unit.--Reuqr (talk) 22:11, 18 May 2020 (UTC)

My feeling is against referring to it as a metric unit in this article. Just because metric prefixes are indiscriminately affixed to quantities does not qualify all those as metric units. Is the dollar a metric unit (kilodollar, megadollar)? Is the dot a metric unit (microdot)? Sorry, that last was a bit silly, but I think you might get the point. Even the dalton, which the SI officially acknowledged in the 8th SI Brochure as taking metric prefixes, is not generally (to my knowledge) regarded as a "metric unit". I see little benefit in stretching a category of "metric units" in this article. —Quondum 22:33, 18 May 2020 (UTC)

If we're not sure, and in the absence of a reliable source (confirming g's status as a metric unit), we should err on the side of caution and remove this unit from the list of metric units. Dondervogel 2 (talk) 22:49, 18 May 2020 (UTC)

Having a section named "Metric units that are not recognised by the SI" makes me uncomfortable. It suggests or even implies that there is a reasonably unambiguously accepted definition of the term "metric unit", which I don't think is true. Such a debate does not belong in this article (but it might belong in Metric system). I would be happier if we just said what we were talking about, e.g. "Units of metric systems that were at some point reasonably established" or "units that people have applied metric prefixes to" (youch), or whatever. —Quondum 23:32, 18 May 2020 (UTC)

The following question pertains to Dondervogel 2's original narrower objection, but answering it would help me in thinking about how we could rewrite this subsection in a way that would also address Quondum's more general objection. The question is this: what about the standard atmosphere, which is also listed in this subsection? Is it less objectionable to call the standard atmosphere a metric unit than it is to call g a metric unit, or is it just as bad, and, in either case, why?

(While I don't expect the following to be decisive as far as the present discussion, for what it's worth, here is one more or less reputable source that explicitly calls ‘milli-g’ a metric unit. It also calls it an Imperial unit, but, as I said above, a unit like g can be both. Anyway, the source is the National Elevator Industry, Inc. Performance Standards Committee, Attachment #11 on page 59 of this pdf.) --Reuqr (talk) 11:43, 26 May 2020 (UTC)

My immediate reaction is that the atm is on no firmer ground than is g as a metric unit. Inclusion in a table of units as recognized for use with SI units does not imply in any way that these were/are metric units. g might at least have in its favour that in some gravitational system it could be formulated as a derived unit. (Can it? My perfunctory mental attempts are not succeeding in even this.) My premise is that a metric unit is one that forms/formed part of a metric system of units at some point. Both of these can be argued to have been used as units (without the qualifier metric). There is precedent in CGS of simultaneously including mutually incoherent units for the same quantity (for energy), so closer investigation would be needed. Incidentally, an isolated reputable source offhandedly referring to something as a metric unit does not make it so for our purposes. We really need something for which the focus is such a classification/definition, or else a demonstration of general acceptance (multiple sources routinely referring to something in the given way). —Quondum 13:34, 26 May 2020 (UTC)

I would describe the

technical atmosphere, defined as 1 kgf/cm², as a metric unit. Dondervogel 2 (talk

) 14:18, 26 May 2020 (UTC)

I am much more comfortable with the

at being called a metric unit. Since it would be pointless to try to make the list comprehensive, the simplest is naturally to remove all units that are in the least controversial in this sense. —Quondum

14:48, 26 May 2020 (UTC)

One question: why is the following sufficient condition in any way objectionable, though? If a unit is defined by the CGPM, then that unit is a metric unit. This condition is not satisfied by e.g. the dalton (because the dalton is not defined by the CGPM, which merely said that it is acceptable for use with the SI). But the condition is definitely satisfied by the standard atmosphere. Resolution 4 of the 10th CGPM (1954, so before the establishment of the SI) reads, The 10th Conférence Générale des Poids et Mesures (CGPM), having noted that the definition of the standard atmosphere given by the 9th CGPM when defining the International Temperature Scale led some physicists to believe that this definition of the standard atmosphere was valid only for accurate work in thermometry, declares that it adopts, for general use, the definition: 1 standard atmosphere = 1013250 dynes per square centimetre, i.e., 101325 newtons per square metre. Moreover, that unit is used in official publications by the CGPM, for example in Comptes Rendus Des Séances De La Treizième CGPM (Paris, 1967 and 1968) [Reports of the Sessions of the 13th CGPM], e.g. on p. A17 (Sauf pour les points triples et pour un point de l'hydrogène en équilibre (17042 K) les valeurs assignées des températures le sont pour des états d'équilibre sous la pression p₀ = 1 atmosphère normale (101325 N/m²) [Except for the triple points and for a point of hydrogen in equilibrium (17042 K) the rated values of the temperatures are for equilibrium states under pressure p₀ = 1 standard atmosphere (101325 N/m²)]. Similar usage is repeated on p. A8 of the Reports of the 15th Session. --Reuqr (talk) 17:40, 26 May 2020 (UTC)

The question is whether we are being prescriptive or descriptive in defining "a metric unit". For example, the

The given "sufficient condition" is not a valid condition. The CGPM does define the dalton ("The dalton (Da) and the unified atomic mass unit (u) are alternative names (and symbols) for the same unit, equal to 1/12 of the mass of a free carbon 12 atom, at rest and in its ground state."). It is not sufficient for the CGPM to define something as a unit for it to be a metric unit; the relevant sufficient condition is for the CGPM to declare it as being an SI unit (though if it described it as a metric unit, would probably also count). The CGPM may define units for various reasons, such as to make pronouncements about their usability or unusability, or more generally, for informative purposes. That the CGPM chose to define the dalton, albeit in a footnote, was presumably to preempt any failure of coordination by other standards bodies. —Quondum 18:38, 26 May 2020 (UTC)

Sorry, I should have been more precise. What I meant by 'define' was something that includes a 'creative act', which I think comes down to the requirement that the definition should involve some sort of substantially arbitrary choice. Let me try to more precisely explain this notion, which I will call sharpen-define to avoid confusion with other meanings of the word define.

Let's say that an authority sharpen-defined a magnitude of a unit just in case the authority specified a numerically exact value for the unit (as opposed to declaring that the unit should be determined by measurement, so that it always has an uncertainty), where the choice of this exact value was substantially arbitrary. By 'substantially arbitrary' I mean that, to the precision of measurement available at the time, there were many other choices for the exact value that could have been made without there being any practical difference.

For example, the CGPM did not sharpen-define the dalton/a.m.u. It did have a choice e.g. to pick a different nuclide, but a different choice in that regard would have made a difference (as far as the magnitude of the unit) that is well above the limits of precision of measurement. On the other hand, once the (ground-state) nuclide was picked, there were no other arbitrary choices, and the unit simply has to be measured.

In contrast, the CGPM did sharpen-define the standard atmosphere. To begin with, even the original historical choice of 76 cm of mercury was arbitrary; 75.99 cm or 76.01 cm would have been just as good. Later, the CGPM just picked an exact value for dynes per square centimetre; they could just as well have picked a slightly different exact value, e.g. 1013250.1 rather than 1013250, without there being any practical difference. But they picked the latter, and the standard atmosphere has the value it has because of that arbitrary choice on the part of the CGPM.

Now, given that CGPM sharpen-defined the standard atmosphere and declared that it adopts it for general use, I just don't see how one can interpret this as merely having been done for informational purposes.

In any event, I think I have a source that is reputable enough to support the claim that the standard atmosphere is a metric unit: A Dictionary of Weights, Measures, and Units, published by Oxford University Press.^[1] Its entry for the standard atmosphere (see here) begins as follows: standard atmosphere physics SI, Metric-c.g.s. and -m.k.s. (This Dictionary lists, for many units, what system of units they belong to, e.g. slug (gee pound, g pound) mass BI-f.p.s., see here.)

(Archon 2488, I certainly agree that deprecated metric units are still metric units. This why I said sufficient condition. I agree that it is not a necessary condition, and nothing in what I say assumes that it is a necessary condition.) --Reuqr (talk) 00:09, 27 May 2020 (UTC)

The CGPM defined the standard acceleration due to gravity and the standard atmosphere as quantities, not as units. —Quondum 02:14, 27 May 2020 (UTC)

I can see how one could say that for the standard acceleration due to gravity, but I don't see how that can be said about the standard atmosphere. To begin with, in the resolution itself we have '1 standard atmosphere = 1013250 dynes per square centimetre'. The usage '1 standard atmosphere =' is exactly how one would write an expression of '{magnitude} [unit] ='. If the CGPM didn't mean to imply that the standard atmosphere is a unit, it is extremely hard to believe they would write it like that, given that they weigh every word in the texts of their resolutions. Next, in their other documents, they certainly use it as they would a unit, e.g. 'pressure p₀ = 1 standard atmosphere (101325 N/m²)' in that Comptes Rendus I cited above. This makes sense, given that using the standard atmosphere as a unit has a long tradition, see e.g. this example from 1899. Finally, the first three editions of the SI brochure (1st, 2nd, and 3rd, appearing on p. 15 of the first two and on p. 16 of the third) list the standard atmosphere in Table 10, 'Units to be used with the International System for a limited time' (see the English translation of the 1st edition, here, p. 15). The entry in the table reads 'Name: standard atmosphere^(e); Symbol: atm; Value in SI units: 1 atm = 101325 Pa'. The footnote says just this: '^(e)Resolution 4 of the 10th CGPM (1954)', which is the resolution I quoted above.

Can I assume we are now in agreement that the standard atmosphere, at least, is a metric unit? --Reuqr (talk) 13:05, 27 May 2020 (UTC)

No, you would be incorrect if you were to assume my agreement. The SI brochures are persuasive that the CGPM considered the atm (and many others) as units, so I would agree that the standard atmosphere can be considered to be a unit. But then we get back to a statement I made above: "Inclusion in a table of units as recognized for use with SI units does not imply in any way that these were/are metric units." I suppose next you're going to want to list the hour, minute, arcdegree, arcsecond, arcminute, nautical mile and knot as metric units. And for all I know, the foot, inch, yard, etc., which are defined in terms of the SI metre, maybe even by the CGPM. —Quondum 16:20, 27 May 2020 (UTC)

I meant on the basis of the cumulative evidence so far. I pointed out the inclusion of the standard atmosphere in the table just to confirm that the CGPM indeed considers it a unit.

I will summarize this cumulative evidence below, but first I need to clear up possible misunderstandings. The other units you mentioned were not sharpen-defined by the CGPM. _{The hour, minute, etc. have been precisely defined in terms of the second by long tradition; similarly, long tradition has defined the angular units. The nautical mile was sharpen-defined in terms of the metre by the First International Extraordinary Hydrographic Conference in 1929. The foot, inch, pound, etc. were sharpen-defined in terms of the metre and the kilogram by the International Yard and Pound Agreement; there are no resolutions of the CGPM that deal with the US/Imperial units. (The Agreement was motivated by the proposal that the metre be redefined in terms of wavelength of light, and was reached by the CGPM representatives of the six English-speaking nations as a sort of a side-project. The CGPM as such was just a background and had nothing to do with that agreement.)}

I emphasize the importance of the concept I called 'sharpen-define' because when CGPM sharpen-defines a unit, then that unit has the exact value it has only because the CGPM—the supreme metric body—declared in an official resolution that this is the value that the unit shall have. (Remember, part of the concept is that the exact value could have been picked slightly differently without there being any practical difference, to within the precision of measurement available at the time; but as metrology advances, that arbitrary choice would start to have measurable consequences). This makes the unit directly dependent on the authority of the CGPM in a way that is qualitatively different from the case when the CGPM merely recognizes an already existing unit, such as what happened with the dalton.

All I ask from others at this point is to reflect on the concept of 'sharpen-defining' and admit that it is at least a relevant consideration, a piece of evidence to be considered in conjunction with other evidence.

And in this case, it is not the only piece of evidence. Another piece of evidence is Ref. ^[1], which explicitly says that the standard atmosphere is a metric unit. I am not saying this book is an infallible Bible, but surely its weight as evidence is not zero. After all, it is a book published by a well-respected academic publisher, and the whole book is dedicated to scientific units of measurement, and nothing else.

Given these two pieces of evidence, I would insist that the burden of proof shifts to those who would claim that the standard atmosphere is not a metric unit. So, is there any actual positive reason (as opposed to a purely skeptical argument) to doubt that it is? --Reuqr (talk) 17:47, 27 May 2020 (UTC)

You claim that the reference (Fenna 2002) "explicitly says that the standard atmosphere is a metric unit". I have reviewed the entry in that reference for "standard atmosphere", where I do not see such a statement. I see "now SI-deprecated as a unit of measure", "Now the derived (non-coherent) unit is defined as 101.325 kPa (14.695 95~ p.s.i.)." and "The subsequent definition by the 9th CGPM of 1948 as a simple derived unit (1 013 250 dyn⋅cm⁻²) and its re-statement in m.k.s. and the SI were only a re-expression of the implied value of the 1941 definition". —Quondum 18:30, 27 May 2020 (UTC)

The term "metric", as applied to units of measure, is a slippery concept in English, because by tradition it has been applied to some units which in the narrower sense are metric, some which aren't, some which are only in some version of the CGS system, and some which could be or not be depending on how you qualified them: A unit of this latter kind is the "metric ton": as a unit of mass it is just another name for the megagram, which is metric; as a unit of force it is the megagram-force which is not metric in the strictest sense of the word as it is not equal to a power of 10 times any product of metric system base units. A unit called "metric" but which isn't in the strictest sense of the word is the so-called metric horsepower, which is (or used to be) equal to 75 kg'.m/s where kg' is one possible symbol for the kilogram-force. When I was in (French language) grade school some 60 years ago "le cheval-vapeur", as we called it, was recognized as a "practical unit" which was not part of the metric system sensu stricto even though its definition was based on metric units and a constant of nature. (For 1 hp we would have said "un cheval-vapeur britannique" but we hardly ever used that unit.) In a similar but not necessarily identical way, the minute, hour and IIRC the day are accepted by the CGPM for use together with metric units but are not strictly speaking metric themselves, and insofar as with very few exceptions only units recognized by the CGPM are legal in France I could add the week and year (whose use is legal in France; the "month" is a unit of time but I hesitate to name it here as its length can vary from 28 to 31 days depending on when it begins).

In the widest sense of the word, of course even Imperial units are "metric units" nowadays: for instance the inch is defined as equal to 2.54 10^-2 m exactly, and the foot, yard, fathom, (furlong?) and (statute) mile are integer multiples of it. — Tonymec (talk) 19:42, 27 May 2020 (UTC)

Quondum: I was principally referring to the very heading of the entry, which reads:

standard atmosphere physics SI, Metric-c.g.s. and -m.k.s.

Meanwhile, I found another source, which says this:^[2]

Atmosphere: A metric unit of measurement of pressure equal to 1.013250×10⁶ dynes/cm² or 1.013250×10⁵ pascals, which is the air pressure measured at mean sea level. It has a dimension of unit of force per unit of area. Used to denote the pressure of gases, vapors, and liquids. Also called Atm, Standard Atmosphere, and Std Atm.

Tonymec: Indeed, like almost all interesting concepts outside of pure mathematics and logic, the concept of 'metric unit' is not perfectly sharp. The sense in which we are, I think, trying to use it here can be at least circumscribed as follows. We all agree that it certainly encompasses all the units belonging to any of the following comprehensive systems: the SI, CGS, MTS, and QES. We also all agree that it definitely does not include the inch, pound, etc. (The users of the Imperial and U.S. Customary systems of units are certainly free to base their systems on whatever they want, including on metric units; that is their business.) The (gram-)calorie is (in my opinion, at least) the prime example of a unit that is definitely metric event though it is not a decimal multiple or submultiple of a coherent unit in any of SI, CGS, MTS, or QES systems. Here is why I think so, in small print. _{The following is largely taken from the open-access Ref. ^[3].The original idea for its definition (which likely predates the actual name for the unit) is metric through and through: take water (much beloved by the early metric system; remember, the original gram was a derived unit, with the density of water being 1) in the amount of one metric unit (cm³) of volume, and raise its temperature by one metric degree; the amount of heat that was required to do this is a calorie. In the 1890s, the British Association for Advancement of Science (BAAS) defined the dyne and erg and recommended the adoption of the CGS system. At the time, electric units (which were the primary object of interest for the BAAS) were in a state of flux, and there was a distinction being made between 'theoretical' and 'practical' units (so that e.g. the volt and the ampere were practical units, while he theoretical units were either stat- or ab- versions of them (e.g. the statvolt and abvolt), depending on whether one used the CGS-ESU or CGS-EMU system). In an 1896 report, the BAAS said that 'There is fairly general agreement in the view that as a practical unit the heat required to raise 1 gramme of water 1°C at some fixed temperature must be taken', recommended using the joule as the theoretical unit of heat, and advised that a 10°C thermal unit based on the heating of 1 g of water should equal 4.2 joule, saying that ‘this secondary thermal unit may be called a Calorie’.} --Reuqr (talk) 21:59, 27 May 2020 (UTC)

This is going somewhere slowly

We will never agree on whether a standard atmosphere is a metric unit unless we first agree on what is meant by "

For what it's worth, I don't think this is how it normally works (i.e. a general definition first, which then settles all controversial cases); rather, I think the discussion on individual controversial cases and the discussion on the general definition go on (and should go on) in parallel and inform each other.

More to the point, however: sure, there is no consensus on the general definition of metric unit, but there is plenty of consensus on that status for many individual units. In the particular case of the standard atmosphere, we now know of two reputable sources that say that the standard atmosphere is a metric unit (click on the link in the reference to see the relevant page on google books): Ref. [1] and Ref. ^[2].

_{(In the first one, look at the heading: ‘standard atmosphere physics SI, Metric-c.g.s. and -m.k.s.’. (Yes, I would agree that the ‘SI’ label is out of place there. But even if one thinks that therefore this source should be given strictly zero weight—arguably an overreaction, but OK—there is still the second source.) The second one is as explicit as it can be: ‘Atmosphere: A metric unit of measurement of pressure ... Also called Atm, Standard Atmosphere, and Std Atm.’)}

I do agree that we should have consensus, but consensus is built one person at a time. Therefore, it would be very helpful if everyone explained why they are not persuaded by the available evidence such as these two sources. --Reuqr (talk) 10:20, 28 May 2020 (UTC)

As you point out yourself, the first source discredits itself by calling the standard atmosphere an SI unit. The second one is more credible, and worthy of consideration, but on its own is it enough? In my interpretation both Fenna and Massey made a simple mistake. Dondervogel 2 (talk) 10:53, 28 May 2020 (UTC)

1. Both here and in the article on the metric units (at some point, I might join the interesting discussion there, too), I think there is a place for a category whose content (though we need to find a better name) is 'units claimed to be metric by only a limited number of sources', and possibly even for a category such as 'units claimed to be metric by some sources, and claimed to be non-metric by other sources'.

2. But why do you find it so implausible that the standard atmosphere might be a metric unit? So implausible, in fact, that you think that a reputable source probably made an error when it said it was one?

2. On its own is it enough? Well, I guess that depends on one's priors. If one thinks that the claim that the standard atmosphere is a metric unit is implausible on the face of it, then one would demand multiple sources; conversely, if the claim sounds a priori plausible, then one reputable source is enough. I find the claim very plausible, and I explained above in some detail why (it has to do with what I've called 'sharpen-defining'). It seems to me that the other editors, however, find the claim a priori implausible. If that is really so, I think it would be helpful if they explained why. And if it is not so, if they don't find the claim particularly implausible, then I do think one and a half reputable source (as I said, I don't think Fenna is completely worthless) is enough. --Reuqr (talk) 11:25, 28 May 2020 (UTC)

The reason it seemed implausible to me that the standard atmosphere is a "metric unit" was because I could see no link with an established metric unit. It seems less implausible now that I see a possible link with the millimetre of mercury, which in turn is closely related to the metre. If the torr is a metric unit (and I think it is), then perhaps 760 torr also is (here I am less certain). Whether a unit is defined by the CGPM (or not) seems irrelevant to me, unless they state it as part of the SI. Dondervogel 2 (talk) 11:43, 28 May 2020 (UTC)

Exactly, CGPM has the authority only to define a unit as being part of SI or formally accepted for use with SI. It has no authority to define what the word "metric" means in common English usage, or what anyone is allowed to call a "metric unit" (not least because there is a lot of historical clutter there that is incompatible with SI); ultimately the difference here is that "SI" is a formal, technical term with a precise definition, and "metric" is an informal term which is much broader in actual use. I don't think it can meaningfully be described as "an error" to use an informal word with a very slightly unusual definition, e.g. accepting the atm as "metric" – because the word ultimately means, I posit, either "deriving [in some not rigorously specified sense, hence allowing for minor disagreements] from formal metric standards" or the slightly circular "a unit that would commonly be referred to as metric".

Ultimately, usage determines the meaning of informal words. If the common definitions of "foot" and "inch" were changed to round values in metric, people might speak about "metric feet and inches" – in the same way you might think of 101.325 kPa as a "metric atmosphere", where "metric" in practice means something like "[now] defined in terms of metric units [where it was not previously]". Thus imperial feet and inches, although now formally defined by the metre, are never called metric, because their physical sizes predate the metric system – and this historical/cultural context plays a large role in determining how people use the word "metric" (or pretty much any other). Archon 2488 (talk) 12:09, 28 May 2020 (UTC)

How is this proposal not a solution in search of a problem? Arguing about the definitions of words is very rarely constructive, and I see no practical benefit in arbitrarily deciding that a given unit in the "metric grey area" is either a "real" metric unit or not. In the case of specific units, I'd suggest following general real-world practice about whether it's a metric unit. By the criterion of common use, the inch is clearly not a metric unit (since although it is defined in metric terms, I have never encountered anyone who would call it metric), the kilogram clearly is, and the calorie and metric horsepower / PS are murkier. Clearly, they are not metric in the sense of being related to base units by dimensional composition and scaling; the calorie being defined as 4.184 J makes it, in an abstract sense, no more metric than the inch being defined as 0.0254 m. But of course, mathematical arguments like these do not explain everything about common use; the reason the various units called calories would be called metric is historical – they were intended to be used with other metric units in a time (i.e. pre-SI) before coherent composition from base units was practiced. If there is a real ambiguity about whether reliable sources, taken as a whole, consider it appropriate to refer to a particular unit as metric, I don't see a NPOV alternative to just repeating that. I don't think you can define all the grey areas intrinsic in the use of normal, informal language out of existence. Archon 2488 (talk) 11:47, 28 May 2020 (UTC)

Which proposal are you referring to? Dondervogel 2 (talk) 12:00, 28 May 2020 (UTC)

I'm sorry, I see that nothing has been explicitly proposed (which is kinda concerning in its own right, as it is not clear what could come out of this discussion) – I was referring to the implicit suggestion that we could "pronounce" whether g, calories, PS etc. are "metric units" for WP's purpose. I suggest that is unworkable and undesirable. I don't see what other kind of conclusion this discussion could come to? Archon 2488 (talk) 12:12, 28 May 2020 (UTC)

Just a quick note about CGPM has the authority only to define a unit as being part of SI or formally accepted for use with SI. This is not so, as evident from the fact that the CGPM predates the SI by many decades. The authority of the CGPM is set by the Metre Convention, which 'established a permanent organizational structure for member governments to act in common accord on all matters relating to units of measurement' (see here). That structure consists of the CGPM, CIPM, and BIPM. Maintaining the SI is merely a tool through which these bodies have since 1960 carried out one of their duties. (There are ten Consultative Committees of the CIPM; only one of them, the Consultative Committee for Units (CCU), is dedicated to units as such.)

Whether a unit is defined by the CGPM (or not) seems irrelevant to me, unless they state it as part of the SI. Given the history of the CGPM (of which the history of the SI is only a part), I find this very surprising. The CGPM has been the supreme authority on all matters metric since it was established in 1875 by the Metre Convention. --Reuqr (talk) 13:00, 28 May 2020 (UTC)

@Archon 2488: OK, now I understand your question. This thread started because of an assertion in the article that the standard gravity is a metric unit. I was just questioning that assertion. Now the scope of the discussion has expanded to include the standard atmosphere. Dondervogel 2 (talk) 13:08, 28 May 2020 (UTC)

I assume CGS would have been similarly formally defined by CGPM when it was the metric system; I also assume that it no longer is, since units such as the dyne are formally deprecated. Since for our purposes in the 21st century, "the formal, modern metric system" and "SI" are synonymous, it doesn't seem relevant to me. I suspect a lot of people here (much less among general readership) would struggle to express great interest in the internal affairs of standards committees, and they don't typically use words exclusively in ways that make sense in the context of formal standards.

The point is that being recognised as the authority to accept, promulgate, and maintain formal definitions doesn't translate into a wider cultural authority to decide how people should use units, or words related to them, in every context of their lives. CGPM can say that the dyne is not recognised as part of SI or formally accepted for use with SI – but this does not translate into a widespread perception that the dyne is not "a metric unit" in the informal sense. Archon 2488 (talk) 13:30, 28 May 2020 (UTC)

I agree with your main claims as far as they go (though your history is a bit off; but I understand that's not the point—for most part). But let me be clear: I have never suggested that the introduction of the SI system has somehow made the dyne, erg, etc. non-metric. Nor have I ever suggested that a unit cannot be a considered a metric unit unless the CGPM says so. _{In case anyone is curious about what, then, I have done, it is this: I proposed a sufficient (but not necessary) condition for something to be considered a metric unit: if the CGPM redefines a unit by arbitrarily setting its value (when expressed in terms of other metric units) to an exact number, then that unit is definitely metric. On the other hand, just because the CGPM didn't redefine a unit this way doesn't mean it is not a metric unit. Also, all bets are off if someone other than the CGPM redefines a unit this way. For example, just because the English-speaking countries redefined the yard and the pound in terms of exact values when expressed in metric units does not mean that the yard and the pound are metric. This recognizes the special status that the CGPM has as far as metric units.}

Another thing I keep saying is that not only did the CGPM define the standard atmosphere in terms of a somewhat arbitrary exact number of newtons per square metre (101325.1 would have worked just as well as 101325), but also there are two (at worst one, realistically 'one and a half') reputable sources that explicitly say that the standard atmosphere is a metric unit. As Dondervogel 2 says, the history of that unit is also metric: 1. it was at one point arbitrarily decided to base it on a round number of centimeters, and 2. this was done in the 19th century in the by-then pretty firmly metric environment of Continental Europe. Finally, I don't think anyone has offered any defeating evidence, i.e. any positive evidence for the conclusion that the standard atmosphere is not a metric unit. Given all this... why, again, can't we agree that the standard atmosphere is a metric unit? (If defeating evidence appears, we'll of course revisit; but there isn't any now!)--Reuqr (talk) 16:02, 28 May 2020 (UTC)

What is special about units defined exactly in terms of metric units? There are many, including the inch, foot, mile, pound, fathom, stone. But two non-imperial examples that that spring to mind are the nautical mile (1852 m) and the astronomical unit (149597870700 m). Those definitions do not IMO automatically qualify them as "metric units". Dondervogel 2 (talk) 16:36, 28 May 2020 (UTC)

I don't get the insistence on this. I, for one, do not even remotely accept the "sharp-define" criterion as being appropriate. You also seem to be over-eager to interpret something as "explicitly mentioned" as shown by the first reference; which only indicates that it is a concept in the context of the SI, Metric-CGS and Metric-MKS systems. Also, when you have to scour the internet to find isolated supporting uses, maybe it is not something we should be putting into an encyclopaedia? I'm kind of fuzzy on why this section is even in this article; it is better covered in

WP:SYNTH, which is exactly what reasoning above is. An encyclopaedia is not intended to define a fuzzy/arbitrary concept and then propound on the details of that fuzzy concept, which is exactly what is happening here. IMO, the article would be cleaner without the section altogether. —Quondum

16:59, 28 May 2020 (UTC)

I found some examples claiming the standard atmosphere is not a metric unit [1] [2] [3]. Dondervogel 2 (talk)

So if the CGPM formally accepted the International Yard and Pound Agreement definition of imperial units, effectively giving their blessing to the UK/US redefinition of those units in metric terms, that would make them metric? As Dondervogel asks, does this apply to internationally used non-metric units like the nautical mile, a nominal rounded value in metres corresponding roughly to a minute of arc? I suspect the real reason these units are not called metric is because their history is largely independent of the metric system. Archon 2488 (talk) 18:56, 28 May 2020 (UTC)

Quondum: 1. I looked more carefully at all the entries in Fenna that are labeled as Metric, and I concede the point. Some of the entries labeled Metric are indeed for headings that are definitely not units, and so that label can be used, in that book, just for indicating context.

2. When you have to scour the internet... No, I think a reputable reference is a reputable reference. As we all agree, it is normally not that important whether a unit like the standard atmosphere is to be considered metric or not, so it's not that surprising some digging might be required. At the same time, as I've already said, I would favor a section (probably indeed in the 'Metric units' article, to which then we would link from the SI article) about units whose status as metric units is controversial. In particular, now that Dondervogel 2 has found actual sources that say that atm is not a metric unit, there should be a subsection for units for which some sources claim to be metric, while other sources claim not to be. (Since we have both kinds of sources.) _{See 'A problem with The Complete Idiot's Guide to Chemistry and the Infoplease references', below.}

3. I'm kind of fuzzy on why this section is even in this article That's fine. I have no objection in principle to delegating some of the work from the SI article to another article. (In fact, I recently created a whole new article basically just for that purpose...) I know you have voiced this opinion before, but at the time I wanted to get some clarity on the immediate question at hand, before I engaged in that larger proposal. Now: it seems that there might be a reorganization of the 'Metric units' article coming up, so I propose we wait until the basic outline of that becomes clearer? You might have noticed that Dondervogel 2 has already removed the 'g as a unit' and the 'standard atmosphere' from that subsection in the SI article, so I assume that subsection is not urgently needing a move?

4. In any event, we cannot make inferences here... No argument there. However, two things. 1. sometimes what one person thinks is inference, another person thinks is obvious, and this can be fleshed out only through discussion (initially both parties suspect the other party is misunderstanding them). 2. Reputable references aren't enough, nor should they be. The editors should also believe that these references are in fact probably correct. (Otherwise, why didn't y'all concede as soon as I found my second reference, which we all agree is a legitimate one?) My 'proposal' was aimed at the other editors (who, to my surprise, didn't think it was simply obvious that the standard atmosphere was a metric unit); for example, I would never suggest that this 'proposal' of mine be incorporated into the article itself. The 'official' justification for saying that the atm was a metric unit was always going to be a reputable source (once it turned out other editors truly didn't think it was obvious). But for that to happen, I would need to convince the other editors that what the reference says is probably correct. It is only at that level that my 'proposal' was supposed to work. (Another function was to let the other editors know where my priors are coming from.)

Dondervogel 2: 1. Ah! That is interesting! Finally, an actual positive piece of evidence that the atmosphere is not a metric unit. (Though I can't find where that is mentioned in the NIST one; can you give a page number?) So now, if we are including the standad atmosphere in any list of metric units, it has to come under a heading 'Units for which some sources say they are metric, while other sources say they are not metric'. But with a better title. _{See 'A problem with The Complete Idiot's Guide to Chemistry and the Infoplease references', below.}

2. What is special about units defined exactly... This is related to Archon 2488's comment. Look, I know by now that none of you like my proposal, but since we are talking about it, let's at least agree on what that proposal is. According to the proposal, it does matter who makes the definition. The CGPM has nothing to do with the inch, foot, mile, etc., and it did not define the nautical mile and the a.u., either. (Archon 2488 asks, 'But what if it had decided to?' I'll answer that below, in my reply to Archon 2488.) As for the general question of why defining something exactly in terms of metric units should matter at all, even if a metric authority is the one doing the defining, let me answer that question with a question. Above, you wrote, It seems less implausible now that I see a possible link with the millimetre of mercury, which in turn is closely related to the metre. That link, of course, is that at some point, the idea was to define the standard atmosphere as 760 mmHg under standard gravity at a standard temperature. Now, why do you feel that this helps the case for the atmosphere being a metric unit? (Would it matter if the proposal were formulated as 30 inches of Hg? What about 762 mmHg? (Note that 30 × 2.54 = 76.2.))

Archon 2488: I promise to answer your question; but I need a bit of a setup. When I was trying to put into words why I felt it was simply obvious that the standard atmosphere is a metric unit, I came up with at least two superficially different reasons that are in fact related. One of them resulted in my infamous 'proposal'. The other one was more historical in outlook, presumably not dissimilar to what you yourself have in mind. I chose to present the first one because it is quite a bit shorter to state, and also because I realized it actually sort of includes (in the case of the standard atmosphere, at least) the historical reasons in it. This is because the CGPM is a very deliberative body. Its resolutions are not arrived at lightly, and are normally products of many years of development, decades even. So let's consider a hypothetical alternative universe in which, say in 1950 (well before the International Yard and Pound Agreement), the CGPM decided to define the yard as a certain exact number of metres. That is a universe in which many other things are different as well (after all, in our universe, it didn't even occur to the CGPM to make such a resolution). For example, it could be that, in the other universe, the CGPM is a capricious body that has a tendency to fire off silly resolutions. Well, in that universe, my 'proposal' doesn't hold, for my proposal deals with a careful, deliberative CGPM. But what if the CGPM is as deliberative and careful in the other universe as it is in ours? Well, this means that, in the other universe, there were sound reasons for the CGPM to make such a resolution, reasons that were absent in our universe. Given these reasons (which we, in this universe, don't know what they are) and the assumed careful nature of the CGPM, it would follow that in that universe it made sense for the CGPM to make such a resolution; but then it is very plausible that these same reasons would be convincing reasons for thinking that the yard is a metric unit after such a declaration. So, yes: in that universe, in which there are some (to us, in this universe, unknown) reasons for the CGPM to make that resolution, in that universe it would follow that the yard had become a metric unit once the CGPM sharpen-defined it.

BTW, if the CGPM simply accepts the International Yard as other bodies have defined it in terms of the metre (like it simply accepted the nautical mile as it was defined by the First International Extraordinary Hydrographic Conference in 1929), that doesn't count, according to my proposal. The essence of my proposal is that the CGPM must carry out a 'creative act', if you will. The 'sharpen-define' is how I spell out what I mean by 'a creative act.' --Reuqr (talk) 23:57, 28 May 2020 (UTC)

On p1 of Gilsinn 1978 I see

"One of the units derived from [the six basic matric units] is the metric unit of pressure, the Pascal (Pa), measured in Newtons per square meter: in basic units, m *kg*s . In addition to the basic metric units and those derived from them, the International Committee of Weights and Measures (CIPM) recommended [6] in 1969 the retention, for the time being, of several other units including the nautical mile, the knot, the bar (l bar = lO^5 Pa) and the standard atmosphere (l atm = 101,325 Pa)."

This extract gave me the impression (late last night) on first reading the standard atmosphere was considered by the author a non-metric unit, but now in the cold light of day I no longer interpret it that way. It certainly does not say so in black and white. I guess I was too tired. Dondervogel 2 (talk) 05:18, 29 May 2020 (UTC)

Regarding your second question, if a standard atmosphere had been defined as 30 inches of mercury, then in my mind it would clearly be a non-metric unit. I was trained as a physicist in the 1970s and I have never consider the standard atmosphere as a metric unit. To be honest I don't really think of it as a unit at all (more like a quantity akin to

standard temperature or standard gravity). It is only because it is 760 mmHg that (and not 762 mmHg) that I am entertaining the possibility. As you can tell I am still far from convinced. Dondervogel 2 (talk

) 05:31, 29 May 2020 (UTC)

And that is fine. This was just to make the point that there is something relevant about defining something in terms of an exact number of metric units, the rounder the better. (BTW, how do you feel about the metric horsepower?)

In my view, the roundness of the numerical value is important—for the metric status of the defined unit—unless it is the CGPM that is making the definition, in which case only exactness matters. My intuition—apparently not shared by anyone else here—is that if the unquestioned principal authority and deliberative body on metric matters goes through the trouble of formally creating, in a resolution, a unit by fiat as opposed to measurement, then that unit by definition deserves to be called a metric unit. (It is possible that there could be other bodies, besides the CGPM, that carry comparable metric authority; for example, in certain instances in the 19th century, the British Association for the Advancement of Science—which formally introduced the CGS system to the world, as well as the units of electricity and magnetism—would probably count as such authority.)

On the larger question of what to do about the 'standard atmosphere', given that we have good references that say it is not a metric unit and at least one good reference that says that it is, I think we should say exactly that: it is a unit of which some references say it is metric while other references say it is not. _{See 'A problem with The Complete Idiot's Guide to Chemistry and the Infoplease references', below.} I agree with Quondum that the article 'Metric Units' is the proper place for such lists. My impression, though, is that the 'Metric Units' article is about to be reorganized, so I think maybe it makes sense to wait until the basic shape of that article becomes clear before largely delegating to it the work of the relevant subsection in the 'SI' article? --Reuqr (talk) 08:30, 29 May 2020 (UTC)

Re metric horsepower, in the absence of a definition of "metric unit", your question hinges on whether the kgf is a metric unit. If the kgf were defined ad 10 N, then it would (in my mind) unquestionably be a metric unit. Given it's definition as g kilograms, I see an argument for (defined in terms of kg) and one against (contaminated by a quantity, g, without a clear metric pedigree). Once one has made this choice (metric or non-metric) for kgf, then my gut feeling is to put the metric horsepower in the same box. I can live with either choice. Dondervogel 2 (talk) 09:31, 29 May 2020 (UTC)

Actually, the thrust of my question was about the role of the non-decimal 75 kg in the definition of the metric horsepower.

As far as the kgf, the whole purpose of the CGPM defining the standard gravity in 1901 was to standardize (and disambiguate) the gravitational metric system (see here). As far as references explicitly claiming the kgf is metric, here are some:

1. 'The kilogram-force and tonne-force are the other metric units commonly used for measurement of force.' (here) Basic Metrology for ISO 9000 Certification, 2012.

2. 'The metric technical unit of force is the kilogram-force' (here). Information Circular, 1981, U.S. Department of the Interior, Bureau of Mines.

3. 'Other metric units. — Other units of some importance are the dyne (dyn), the force unit in the CCS system; the sthene (sn), in the metre-tonne-second system; and the kilogram-force (kgf), or kilopond (kp), the metric technical unit of force' (here) Kempe's Engineer's Year-book, Volume 101, 1996.

4. 'In addition a number of traditional metric units are outside SI, and their use is deprecated. These include the kilogram-force, calorie, micron, torr, fermi, metric carat, stere, and gamma.' (here) AIA Metric Building and Construction Guide, 1979.

5. 'The use of the metric engineering units — kilogram-force per square centimetre…' (here) Metrication- Managing the Industrial Transition, 1975. --Reuqr (talk) 10:59, 29 May 2020 (UTC)

If the metric horsepower were defined as 750 N, I would describe it as a metric unit without batting an eyelid. Dondervogel 2 (talk) 11:02, 29 May 2020 (UTC)

I think we should keep in mind in all this that terms tend to be context-dependent. In the early 1900s, the term "metric unit" by necessity had a broad meaning. In the modern era, it has come to be used in some contexts as a synonym for "SI unit", as it seems what is meant in the case for NIST (e.g. here), which suggests that it might normally mean "the units of the prevailing metric system". A broad trawl is insensitive to these contexts, and tends to pick up the union of all usages. This debate is also context-sensitive: in the context of this article, I think it is out of place, and in the context of

Metric unit

, it seems to be premature.

For some light humour about the reliability of reputable sources (spot the mistake), I happened across this in Fenna:

light-microsecond astronomy The distance travelled by light in 1 μs, = 299.792 458 mm (11.803~ in). Not much used, but the natural unit that would have better served as the metric unit of length (see metre).

light-nanosecond astronomy. Symbol lns. The distance travelled by light in 1 ns, = 0.299 792458 mm (0.011 803~ in).

—Quondum 12:45, 29 May 2020 (UTC)

We should count our blessings that there actually isn't anything like long and short scales in this context… --Reuqr (talk) 15:33, 29 May 2020 (UTC)

Indeed. The SI context goes a long way to removing ambiguity, as long as one stays strictly within it. Though SI is not devoid of similar issues, such as the quantity mess with angle vs. rotation and level vs. gain (as evident from some not-so-agreeable back-and-forth in metrology circles and the sudden disappearance of some of these quantities from ISO 80000), and many particle physicists still seem to refuse to acknowledge that F = q₁q₂/r² is not the only way to define charge, leading to lots of ostensibly reliable sources with quantity definitions and formulae that are inconsistent with those of other sources. —Quondum 16:36, 29 May 2020 (UTC)

A problem with ''The Complete Idiot's Guide to Chemistry and the ''Infoplease references: Appolgies, I should have cought this when Dondervogel 2 first brought these two up.

Point 1: These two are in fact one and the same. Not only are the texts, as best as I can tell, word-by-word the same, but also the parent directory of that Infoplease page (above the heading Chemistry: Important Terms and Units, there is something that looks like this: 'Home > Math & Science > Chemistry > Chemistry: Important Terms and Units'; click on 'Chemistry') literally says The Complete Idiot's Guide to Chemistry (see here). And on its table of constents page it says (towards the end of the page), 'Excerpted from The Complete Idiot's Guide to Chemistry © 2003 by Ian Guch'.

Point 2: If you search for 'metric' in the Idiot's Guide, on p. 4 you will find this: Because chemistry is used throughout the world, it uses the International System of units (called the SI system, from the original French “Système international d'unités"). The International System is just a fancy name for the “metric system,” so you will probably see both used interchangeably. (See here.) Therefore, when the Idiot's Guide, later in the book, says though it's not a metric unit when referring to the atmosphere (here), it is reasonable to interpret that as merely saying that the atmosphere is not an SI unit, which no one disputes.

Point 3: It is true that the subsection from Point 2 is missing in the Infoplease coverage, i.e. nowhere on the Infoplease site does it say that The International System is just a fancy name for the “metric system”. However, I hope we are all in agreement that this doesn't matter, once it is established that the Infoplease content is excerpted from the 'Idiot's Guide'.

In light of this, we are back in the situation where we know of no positive evidence for the claim that the 'standard atmosphere' is not a metric unit (in some sense of 'intermediate breadth').

I accept that everyone else here is highly sckeptical about the claim that the 'standard atmosphere' is a metric unit (in some sense of 'intermediate breadth') and that, therefore, a single reference that says that it is a metric unit (this one: Ref. ^[2]), no matter how reputable, is not enough to justify the inclusion of the 'standard atmosphere' in any list of units that are definitely metric.

^{As has been pointed out by multiple people, there is a whole spectrum of meanings of the term 'metric unit'. On the narrow end, 'metric unit'='SI unit'; on the maximally broad end, any unit officially defined in terms of other metric units is itself a metric unit (under this reading, even the yard and the pound are nowadays 'metric' units). Here were are trying to circumscribe a meaning intermediate between these two, which would seem to be the most interesting one for the purposes of a Wikipedia article named 'Metric units' and for the purposes of the section Metric units that are not recognised by the SI of the present article (the content of the latter, we all seem to be in agreement, should eventually somehow be transferred to the 'Metric units' article).} --Reuqr (talk) 20:03, 30 May 2020 (UTC)

I like getting to the nub of the problem, which you have pretty neatly stated in your small print. The discussion of "With what definition(s) of metric unit should we be working?" probably belongs at

Talk:Metric units. —Quondum

20:38, 30 May 2020 (UTC)

I know. But since this is where we've done so much discussion on the 'standard atmosphere', it makes sense to continue discussing that particular unit here, I think?

Anyway, I did more scouring trying to find sources that explicitly claim that the standard atmosphere is not a metric unit (in some sense of 'intermediate breadth'). I found some, but, in my opinion, they do not manage to make that case nearly as strongly as Ref. [2] makes the case that it is a metric unit. But perhaps others will disagree. Anyways, here they are.

1. Northwest Rosarian. On p. 7: In scientific writing, however, water potential is expressed in MegaPascals (MPa), since an Atmosphere is not a metric unit. This is a claim by 'Gary Ritchie, Ph.D., Master Rosarian, Olympia Rose Society'. There is no other discussion of units in his article. For all we know, this is a case where 'metric unit' really means 'SI unit'. In any event, not clear how much authority this source really has.

2. British Columbia Geological Survey Branch Style Guide. On p. 94:

In cases where both SI and Imperial measures are cited, the Imperial measure should appear in parentheses, however, where Imperial measures are used in quoted passages, they are quoted verbatim with the SI equivalent in square brackets. In cases where a quantity was clearly measured originally in Imperial units, it is preferable to use this procedure, for example, a pressure of 2 atmospheres is more meaningful to most readers than a pressure of 202.6 kilopascals and it should therefore be expressed as "a pressure of 2 atmospheres (202.6 kPa) ".

Though only a style guide, this source is more reputable. Nevertheless, it is puzzling that it refers to the atmosphere as an Imperial unit. Whatever one may think of the case for the atmosphere being a metric unit, the case for it being an Imperial unit is surely weaker. The exact values '760 mmHg' and '101325 Pa' are surely more at home in a metric system than the approximate values '29.921 inHg' and '14.696 lb/in²' are in the Imperial system. (My guess is that the author conflates 'Imperial units' with 'familiar units'.)

3. NBS Technical News Bulletin On p. 123: Units not compatible with the SI, nor any other metric system" "calorie; British thermal unit; entropy unit; roentgen; atmosphere; torr; millimeter of mercury. This is definitely a reptable source. However, what they say is that these units are 'not compatible' with any metric system. In my opinion, what this means is 'not coherent' in any metric system. After all, the list includes the roentgen, whose pedigree is metric through and through, with essential references to the esu unit of charge (the statcoulomb). (I personally think the (gram-)calorie is also indisputably metric, and I've explained why in one of my entries above; but someone else here said its status as a metric unit is 'murkier').

4. Geological Survey Professional Paper (U.S. Geological Survey). There is a table with two columns, one labelled 'Metric unit', the other 'Inch-Pound equivalent', and the atmosphere appears in the second column. I am not sure what's with geological surveys in North America, which seem to think that the atmosphere is an inch-pound/Imperial/U.S. customary unit (see 2. above). All I can say is that that doesn't make any sense, as I've already commented above under 2. --Reuqr (talk) 23:55, 30 May 2020 (UTC)

Wherever we debate it, I think it is counterproductive to try to determine whether to classify something as a "metric unit" until you have defined which of many (as yet fuzzy) definitions to use. Even when assessing a source, you need to make the same classification about their use of the term. You are sort of doing this, but only coarsely, in relation to assessing a unit, not really assessing the definition of "metric unit". It feels like debating whether a ball is "small", without defining "small". Many units are traceable for their definition to some point of the metric system's history, the standard gravity and standard atmosphere included, but never were decimal-coherent. Another group are units that were defined as simple multiples of SI units, but are not traceable in the same way (e.g. to the standards bodies), such as the metric carat and metric pound, putting them into a disjoint category. Some were just half-baked non-coherent metric additions that had some status as "practical units", like the calorie and the mmHg. Some were proposals for coherent (or non-coherent) derived unit names that didn't seem to get to acceptance (e.g.

benz). Once these categories have been defined (and we stop clinging to the term "metric unit" as a classification), I'm pretty sure there will be little disagreement or debate. C'mon, you've acknowledged the problem, stop sidestepping it. —Quondum

01:15, 31 May 2020 (UTC)

I don't know about the torr (though this source says it is not). The mmHg is non-controversially metric, and the original definiton of the standard atmosphere was as 76 cm of mercury at standard gravity, so perhaps the mmHg is more relevant to the status of the standard atmosphere. I'll have to do a bit of research about how the split between the Torr and the mmHg actually happened. --Reuqr (talk) 12:00, 28 May 2020 (UTC).

You write above "[i]f the torr is a metric unit (and I think it is), then perhaps 760 torr also is (here I am less certain)" – this seems to be the wrong way around to me. The torr is defined with reference to the standard atmosphere, so surely if that is not a metric unit, the torr cannot be? Archon 2488 (talk) 12:34, 28 May 2020 (UTC)

My reasoning is simple enough: if the torr is (or was originally) equal to 1 mmHg and the mmHg is metric, then the torr is also metric. Dondervogel 2 (talk) 13:05, 28 May 2020 (UTC)

Does anyone know when and how the split between the torr and mmHg actually happened? (In other words: when 760 mmHg became slightly different from 1 standard atmosphere, who decided that 1 Torr should remain 1/760 of the standard atmosphere as opposed to remain 1 mmHg?) I see no mention of the torr (though there are mentions of the mmHg) in the Proceedings of the 10th CGPM, which is where the standard atmosphere was defined (here; unfortunately only available in French, but google translate does a good job with these texts, I found). The Wikipedia article on the torr just states without reference (!) that, following the redefinition of the standard atmosphere by the 10th CGPM, 'the torr was then redefined as 1/760 of one atmosphere'. --Reuqr (talk) 13:13, 28 May 2020 (UTC)

I assume it's because mmHg is ambiguous; g is not constant around the world, so the actual value of the mmHg (or, the height of a column of mercury that a given pressure can support) depends on where it is measured. Defined as 1/760 atm, the torr is effectively defined by the pascal, which removes this problem. Archon 2488 (talk) 13:34, 28 May 2020 (UTC)

To expand: I suspect that what happened with the mmHg was that the empirical definition was "metricated" (i.e. the previous definition was realised then remeasured in some way in terms of SI), and this value in pascals then became the formal definition of the unit. The fact that this was not the same procedure followed for the torr explains why the units do not exactly agree; the relevant empirical values, such as local gravity and pressure and the assumed density of mercury at the ambient temperature, would make it difficult to realise the mmHg as exactly 1/760 atm. Why keep the torr as a separate unit rather than just deprecating it, I have no idea. Archon 2488 (talk) 13:44, 28 May 2020 (UTC)

Archon 2488: In 1948, the 9th CGPM defined the standard atmosphere (principally for the purposes of defining the temperature scale, but, as the next CGPM in 1954 clarified, also for 'general use') as 1013250 dyn/cm² (the pascal hadn't been defined yet), with the following note: 'This pressure corresponds to that exerted by a column of mercury 760 mm high having a specific mass of 13.5951 g/cm³ and subjected to a gravitational attraction of 980.665 dynes/gram. Except for work of the highest precision, it can be assumed that comercially available pure mercury at 0°C has a specific mass of 13.5951 g/cm³ on average in the column.' This same definition already appeared in an annex to the 8th CGPM in 1933 (here, p. 74). The value 980.665 dynes/gram is the standard gravity that the CGPM defined back in 1901. Now, 980.665 × 76 × 13.5951 = 1013250.14; if we assume that the density of mercury has an uncertainty of ±0.5 in the last significant digit, which is ±3.6×10⁻⁶ relative uncertainty, and that the other two numbers are meant to be exact values, that propagates to ±3.7 dyn/cm² uncertainty in the value of the standard atmosphere. The next step is what you call 'metrication', and I call 'sharpen-defining': the CGPM droped the uncertainties and declared that the standard atmosphere shall have the exact value of 1013250 dyn/cm². They could have just as well said 1013250.1 dyn/cm²; any value in the interval 1013250±4 would have been perfectly fine, given the measurement uncertainties of the era.

More on the torr: While looking for references for the kilogram-force (see above), I stumbled upon this one, which (contrary to the reference in my first entry in this subtopic, above) says that the torr is a metric unit: 'The use of the metric engineering units — kilogram-force per square centimetre, kilogram-force per square millimetre, bar, torr, etc. — is discouraged and should be discontinued.' (here) Basic Metrology for ISO 9000 Certification, 2012. Possibly they really mean mmHg and assume they are the same, but who knows.

As far as the history of the torr, Gyllenbok and Jerrard and McNeill say that the torr was adopted by the British Standards Institution in 1958. Gyllenbok specifies that the standard in question is British Standards 2951. Jerrard and McNeill add that the torr 'was used in Germany for several decades previous to this'. The OED gives these quotations:

1949 S. Dushman Sci. Found. Vacuum Technique i. 4 In German literature, 1 Tor (or 1 Torr) is used to designate 1 mm Hg pressure.

1958 Gloss. Terms Vacuum Technol. (B.S.I.) 7 Torr, a unit of pressure defined by the relationship: 760 torr = 1 standard atmosphere (atm) =1013250 dyn/cm² exactly. 1 torr is equal to the conventional barometric millimetre of mercury..within 1 part in 7×10⁶.

The first seven editions of the SI Brochure included the torr in their lists of deprecated units, and they defined it as 1 Torr = 101325/760 Pa. --Reuqr (talk) 22:44, 29 May 2020 (UTC)

If you want an explicit mention of the torr as a metric unit, Fenna (p. 19) says: "A distinct old metric unit for atmospheric pressure was the torr." —Quondum 22:54, 29 May 2020 (UTC)

Ah good catch. I thought Fenna wasn't fine-grained enough to distinguish the torr and the mm Hg, because in its 'head of liquid' entry (p. 122) it said 'The mm of Hg, which is identically the torr, is the usual unit for blood-pressure readings.' But in its entry for the torr (p. 293), it says, 'For gases, particularly the atmosphere; designed to equate with the millimetre of mercury (see 'head of liquid') but to have a precise value for the standard atmosphere, it is thus defined as 101325/760 Pa…' The writing is definitely not the best specimen of clarity I have ever seen, but it seems to say that mm Hg and Torr are not strictly identical after all. --Reuqr (talk) 23:18, 29 May 2020 (UTC)

Yep, it has its imperfections. Interesting to note the phrase "old metric unit", which is suggestive that something might cease to be a "metric unit" at some point, like "estwhile metric unit". —Quondum 23:40, 29 May 2020 (UTC)