Home PageAbout MindatThe Mindat ManualHistory of MindatCopyright StatusWho We AreContact UsAdvertise on Mindat
Donate to MindatCorporate SponsorshipSponsor a PageSponsored PagesMindat AdvertisersAdvertise on Mindat
Learning CenterWhat is a mineral?The most common minerals on earthInformation for EducatorsMindat Articles
Minerals by PropertiesMinerals by ChemistryAdvanced Locality SearchRandom MineralRandom LocalitySearch by minIDLocalities Near MeSearch ArticlesSearch GlossaryMore Search Options
The Mindat ManualAdd a New PhotoRate PhotosLocality Edit ReportCoordinate Completion ReportAdd Glossary Item
Mining CompaniesStatisticsThe ElementsUsersBooks & MagazinesMineral MuseumsMineral Shows & EventsThe Mindat DirectoryDevice Settings
Photo SearchPhoto GalleriesNew Photos TodayNew Photos YesterdayMembers' Photo GalleriesPast Photo of the Day Gallery
cummingtonite vs. anthophyllite
Posted by Tom Mortimer
Frank K. Mazdab November 23, 2018 08:04PMHi Tom,
I realize you're asking somewhat of a rhetorical question. In some instances I imagine the ID is based just on locality; if the geologic environment of a particular locality favors one over the other (for example, in the somewhat uncommon cordierite-anthophyllite rocks), or if previous petrography indicates only one polymorph is present (or most common) at a particular locality, then I suppose, whether appropriate or not, every sample from that locality would be labelled similarly.
I agree that at localities not previously well-characterized or where both polymorphs may likely be present, and especially when the ID is entirely visual and the material is a bit non-descript, caution is warranted. But with amphiboles (and indeed with many other minerals too), there are additional challenges. Even if you're satisfied your amphibole is orthorhombic, is it anthophyllite or is there sufficient Fe for it to be ferro-anthophyllite? Is it anthophyllite or is there sufficient Al for it to be gedrite? Simply based on the surprising number of incorrectly-identified materials I've bought even from reputable dealers, I wouldn't be too shocked that discover that a substantial percentage of photos here across the board are unintentionally mislabeled. The hope of course is that over time these materials get better characterized and then eventually corrected.
Edited 1 time(s). Last edit at 11/23/2018 08:05PM by Frank K. Mazdab.
Alfredo Petrov November 23, 2018 08:09PMTom, I suppose many photos are from localities that have been studied and the literature has reported either one or the other of these and the field collector/photographer just follows that. Or the specimen came from a dealer who had a lot of the material, got one piece analyzed and assumed the rest were the same (usually, but not always, a valid assumption), and the photographer just copied what was on the label. Ditto for specimens derived from old collections. Of course with amphiboles there is the additional complication that species definitions have changed over time and what used to be XYZ-ite might not be that species anymore. Some fuzziness is inevitable in naming amphiboles.
Tom Mortimer November 23, 2018 09:39PMFrank & Alfredo, Thanks for your comments. I agree with your responses regarding the degree of certainty of many mindat photoed species, particularly amphiboles and phosphates. The problem can be compounded when collector A posts photo of mineral X from a locality and collectors B, C, and D copy that identification .
My candidate cummingtonite, Gilsum Road, Surry, NH (photo with EDS attached, 5 mm FOV) was self collected. Gedrite and cordierite are present in this outcrop. The EDS carbon response is due to the carbon coating on the polished epoxy grain block.
An APFU calculated from the EDS analysis atomic percents gives: (Mg3.35Fe1.83Al0.21Na0.46)Σ=6.27(Si8)O3.57, normalized for 8 atoms of Si. Oxygen is low, but frequently have problems with it.
The Handbook of Mineralogy gives a cummingtonite analysis of (Mg4.44Fe2.02Al0.46Ca0.20Na0.18Ti0.03Mn0.03)Σ=7.20(Si7.36Al0.64)Σ=8.00O22(OH)2.5 .
I am planning to submit a sample for PXRD.
Frank K. Mazdab November 23, 2018 10:18PMHi Tom,
there's way more of an issue with that analysis than just the O. The weight % of the elements are clearly wrong... it looks more like those are oxide weights than element weights, for one. I'll attempt to normalize your data and report back.
EDIT: If you assume the original "element wt%" are in fact oxide weights, then the analysis can be normalized to a not-so-great formula, using the normalization protocol 15eK (T+VIM+VIIIM = 15, and which assigns all Na to the VIIIM site). Fe3+ and OH (and H2O) are calculated:
The total comes out to 99.1 wt% (H2O included; but these are based off of the original incorrect oxide weights, so is a bit meaningless).
The normalization routine 15eNK (which assigns all Na to the A-site) also works, and yield a similarly odd formula:
This one calculates a total of 97.8 wt% (H2O included; same additional caveats as above).
Both normalizations approximate to some strange "sodian tetra-ferri-ferri-gedrite"-like material that isn't currently a mineral (note that the listed distribution of Fe2+ and Mg between VIM and VIIIM is purely artificial; you'd need an X-ray structural analysis to verify that).
In any case, the reality, however, is almost certainly not that you have discovered a cool new mineral, but rather that the analysis is terrible... sorry. And the ultimate outcome is that I wouldn't use this analysis to attempt to name this material.
EDIT #2: As an aside, if you look up a mineral like anthophyllite, for example, on webmineral (a website of ancillary use to mindat), it gives you both the typical element wt% values (for example, Si = 28.78 wt%) and the corresponding oxide wt% values (again for Si, SiO2 = 61.56 wt%) for the listed mineral. Your value for "Si" of 40.45 falls almost in the middle of those numbers and so does not reconcile well with either value, even accounting for some minor Al substitution. That's a problem!
Edited 3 time(s). Last edit at 11/23/2018 11:21PM by Frank K. Mazdab.
Frank K. Mazdab November 23, 2018 11:41PMLOL... it is fun! Even if I won the lottery, I'd be spending my days doing these calculations and scratching my head over the results. I'd simply be doing it with better equipment... ;-)
And to get these specimens for study, I might travel to all of the exotic places these minerals occur (and also enjoy the sights, lounge around, and wine&dine at the same time... of course).
Ben Grguric November 24, 2018 08:51AMDo you have access to or know someone with a petrographic microscope? Crushed material will be suitable, you don't necessarily need a thin section.
Anthophyllite is an orthorhombic amphibole. Crystal sections will have straight extinction.
In contrast cummingtonite is monoclinic, and the y:z axis extinction angle can be up to 20 degrees. Most characteristic is the fact the cummingtonite is usually multiply twinned on (100), with narrow twin lamellae.
Alternatively spend $50 and get an XRD analysis.
Tom Mortimer November 24, 2018 03:42PMFrank: The table in my attached pdf lists atomic percents, not weight percent oxides. Oxygen would not be a table entry in a weight percent oxide report. So your assumption that the analysis results "are in fact oxide weights" is incorrect.
My attempt at an APFU computation, based on the EDS element atomic percents, was only to show that the cation to silicon ratio was similar to that given in the Handbook of Mineralogy for cummingtonite. My post in no way remotely suggests that I may have discovered "a new cool mineral." I suffer no such delusion - give me a break! I believe the element composition reported in my (polished grain) analysis is consistent with a cummingtonite-anthophyllite mineral, both in the elements that are present and those that are not.
Ben: I have an amateur friend with a petrographic scope. I will forward him a sample and see what he can determine. Thanks for the suggestion.
I will send off material for an PXRD next week.
Frank K. Mazdab November 24, 2018 08:30PMHi again Tom,
The first column in your EDS spectrum after the element list literally say "weight %". Those are the numbers I used in my normalizations, as I typically start with weight % element, convert to moles, and so on. But you are correct, oxygen would not included in a table of weight % oxides. But no amphibole has anywhere near 45.60 wt% Si, and your values for Fe+Mg are similarly way too high. Of course, the issue, as you already noted, is with your oxygen, and the unfortunate way that standardless EDS happily forces a total to 100%. So a crappy O determination (no one's fault... just a limitation of the instrument) in turn propagates into all the crappy weight % data for the other elements. So then those values have to be ignored (my mistake) and only the atomic % then used instead. That is not how you want to be doing mineral analyses.
Now, it is true that your atomic % ratio of (Mg+Fe+Al):Si of (17.38+20.79+2.51):45.60 is close to 7:8 (closer in fact than your own calculations indicate), and indeed that's the only saving grace here (but if you opt to include the Na, it gets worse; indeed, then it actually normalizes closer to a pyroxene than to an amphibole!). And note that based on your results, if all Fe is presumed to be Fe2+, then your amphibole would be a "ferro-" end-member because the atomic % ratio of Mg:Fe is 17.38:20.79 (column 3; again, I have no idea how you derived your apfu numbers?).
No, I wasn't suggesting you had a new amphibole nor was I implying that you were proposing that. You started the thread asking how posters here differentiate anthophyllite from cummingtonite, and replies from several posters here indicated that chemically, that differentiation probably isn't possible. XRD should certainly be effective, and Ben's suggestion is also a good one, although it should be noted that depending on grain orientation, orthorhombic minerals can at times have inclined extinction, and monoclinic minerals can at times have parallel extinction; this shouldn't be an issue in a grain mount because the grains would not be random cuts and would always be sitting on cleavage surfaces.
Anyway, that probably should have been the end of it. But for minerals ID'ed through chemical analysis, the first step is to get a good chemical analysis. While it may be possible to extract some useful information from the analysis you provided, it is in fact not a good chemical analysis. And as one's unknowns increase in complexity, one's ability to imagine a formula by extracting what appears to be useable data out of a bad analysis becomes increasingly suspect. I guess I should have heeded my own advice and not tried to normalize that jumble of numbers myself... my bad.
Mathieu Butler December 05, 2018 11:42PMHi Ben , Frank
I am the amateur PLM guy Tom mentioned, I analyzed the xl fragment grains mounted on a slide (with oil) and a few mounted on a homemade spindle stage just to be sure - all grains were parallel extinction.
Then Tom's XRD came back confirming it - orthorhombic anthophyllite.
It seems PLMs can still help with polymorph id's instead of paying for an XRD
I'm curious as to how you guys know so much about optical mineralogy? off the top of your head?? : )
I'm self taught and still learning - can you suggest any learning sources (books, web sites, experts on midat)?
I have "optical crystallography" by Bloss and his older spindle stage book
Frank K. Mazdab December 06, 2018 01:06AMHi Matt,
thanks for the update on the microscopy. Certainly the combination of your microscopy work and the independent XRD make an anthophyllite ID unequivocal.
As for my background in transmitted light microscopy, I learned from one of the experts in the field... Robert Hutchinson. He was my optical mineralogy professor one semester and then my petrography professor the next when I was an undergrad at the Colorado School of Mines. He was a hard-ass, and a stickler for absolute precision in making measurements and in keeping organized notes; I think students were a little scared of him (or perhaps scared of what grade they might get in his classes). But he was a powerhouse of knowledge, and probably because of his requirement for excellence, the students in his classes came out with a great education! Tragically, he was killed in a freak car accident a few years ago.
Later when I had to actually teach optical mineralogy and petrography to students, one really learns it then! And I've been in love with the subject ever since. Indeed, in recent years, I've slowed down on collecting hand sample-sized rocks with pretty crystals for my collection, and switched over to ugly billet-sized rocks to make thin sections.
If you'll excuse a bit of shameless self-promotion, I'd recommend my own website for additional knowledge of minerals in thin section, especially if you're already familiar with the common minerals and want to see what some weirder ones look like, or want to see what some less common rock types look like. I have a set of scanned images of thin sections in regular light and under crossed polarizers at:
and I have a large collection of short videos (~30 s each) of minerals in thin section (PPL, XP, and optic figures) at:
As for books, there are several I like: Nesse's Introduction to Optical Mineralogy is great for common minerals, Stoiber & Morse's Microscopic Identification of Crystals includes unusual topics like the kinds of dispersion, and the Handbook of Mineralogy series (now also fully online and updated at the MSA website) includes optical data for over 4000 minerals.
Edited 1 time(s). Last edit at 12/06/2018 01:12AM by Frank K. Mazdab.
Mathieu Butler December 06, 2018 03:01PMThanks Frank, great web site - I think I may have found it through google before but lost track
It is sad that not many (or any?) schools still teach optical mineralogy - except the McCrone institute right?
Where did you come up with the extinction angle to differentiate the cummingtonite / anthophyllite?
I would not have thought of that - maybe because the mindat optical properties do not go that deep
Is there some handy list of optical tricks for the different specific dimorphs or you just have to take it on a case by case basis?
I'm working with fragments / single xls (no access to making thin sections) and trying to master the spindle stage
Digging in deeper
Edited 1 time(s). Last edit at 12/06/2018 03:09PM by Mathieu Butler.
Donald B Peck December 06, 2018 03:56PMMatt,
I am pretty much self taught, also. But I have to thank Curt Segeler, who was very generous with his time and taught me the skills with the spindle stage. I have Don Bloss's Optical Crystallography, and also his book on the Spindle Stage, both of which are very good (and a little difficult). The "optical" text that I like a lot is Optical Mineraogy by Paul Kerr, McGraw Hill. It is comparatively easy to learn from, has good data, and is as good for working with fragments as with thin sections.
Frank K. Mazdab December 06, 2018 07:44PMHi again Matt,
There's no "trick" about extinction angle. It's a property of the mineral, and is ultimately based on how the three principal perpendicular optical directions are super-imposed on the three principal but not necessarily perpendicular crystallographic directions. The optical difference between anthophyllite and cummingtonite is a function of their different crystal symmetries (orthorhombic is higher than monoclinic). It's really pretty elegant, and any of those optical mineralogy textbooks will lead you through the development of those relationships.
Interestingly, in grain mounts, because minerals tend to sit on cleavage surfaces and so cleavages are almost always parallel to the stage, extinction angle relationships probably always (or essentially always) conform to what's expected. In thin section, however, because the possibility exists that minerals can be cut at strange odd angles to both the optical and crystallographic directions, there are occasional exceptions to the general extinction angle expectations as they relate to the underlying mineral symmetry. Higher symmetry minerals can occasionally fortuitously appear lower symmetry, and lower symmetry minerals can occasionally fortuitously appear higher symmetry. This can be a bit disorienting and can lead to the possibility of a misidentification, and hence the recommendation is always to check multiple grains, whenever possible.
Mathieu Butler December 06, 2018 09:43PMHi Frank,
By trick, I just meant that it would be nice if under the cummingtonite mineral data (on mindat ,etc) that they mentioned that the extinction angle was an easy way to tell the difference between it and its dimorph.
And it would be nice if all the dimorphs had handy techniques like this but I guess it's a case by case basis and up to your own knowledge of optical property diffs.
Yes, very handy that the grain cleavages are parallel to the stage - but like you said , it can't hurt to double check (with a grain mounted on a spindle stage in my case)
Thanks again - still checking out your website
Hi Don - I have your handbook also thanks!
Frank K. Mazdab December 06, 2018 10:17PMHi again Matt,
you might find this site useful:
The optical data on over 4000 minerals is pretty complete, and although "extinction angle" is not explicitly stated as such under the main optical properties heading for some given mineral you look up, it can then be determined from the orientation data contained within that section. If you compare the "orientation" data for anthophyllite and cummingtonite I think you'll see what I mean. The overall optical data tend to be more complete and extensive there than what is presented here on mindat (although the visual representation of birefringence here on mindat is useful and a nice touch).
Mathieu Butler December 10, 2018 07:12PMHi Frank / Don
I was given a copy of the older (1959) "optical mineralogy" book by Kerr and under anthophyllite they specifically say:
Anthophyllite resembles tremolite-actinolite and also cummingtonite but may be distinguished from these by it's parallel extinction
and of course the cummigtonite section mentions it has inclined extinction
So, yes, the old texts still do have good data - thx!
Donald B Peck December 11, 2018 03:25AMExtinction is one of the easiest properties to observe, with an elongated mineral crystal or fragment on the stage of a polarizing/petrographic microscope and between crossed polarizers, any fragment (not isotropic) goes extinct every 90o of stage rotation. Anthophylite is extinct when parallel to the cross-hairs. Cummingtonite is at extinction with a maximum angle of 15o to 29o to the cross-hairs ( the angle increases with Mg content; and is at max with the b-axis [or Y] vertical). Actinolite/Tremolite has a maximum extinction angle of 10o to 20o if the b-axis [Y] is vertical.
Many years ago, when I was wondering where my next dollar was coming from, I bought a used A0-60 American Optical scope of the type used in high school biology classes. Its advantage is that the stage can be centered. Anyway, I added a cross-hair reticle to the ocular; a revolving graduated circular stage; a polarizer below the stage (polaroid film), and inserted a Lucite body below the prism optics that takes a slide with the analyzer and a 2nd slide with either wave plates or a quartz wedge. It isn't perfect, and it doesn't have a Barlow lens (that is sort of solved by holding a loupe above the ocular). I now have a homemade spindle stage and oil cell for it. And it was with this that I learned what I know (???) of "optical".
It will do most anything except it isn't very good with optic figures. Another problem is that there is some color in the polaroid film. If you like to tinker, altering an inexpensive scope is fun. It is also fun to use, and with a spindle stage it is quite powerful. Oh, and for extinction all that is needed are the crosshairs, graduated revolving stage. polarized, and analyzer, which can go in/on the ocular. A childs toy scope will work.
I am a dedicated advocate of trying to determine the species name without six figure electronic marvels . . . those are a last resort ;<}
Frank K. Mazdab December 11, 2018 04:54AMHi Donald,
Your homemade petrographic scope design is very clever. It certainly is a pretty economical way to acquire a petrographic microscope. Although I've noticed a few models on Amazon that are amazing cheap (as low as ~$800), I wonder how decent the optics would be on those?
As to the orthorhombic vs monoclinic distinction through extinction angles, without a spindle stage, however, those extinction rules only work most of the time (probably essentially all the time in a grain mount), but there are occasional exceptions in a thin section for some oblique sections of otherwise "parallel extinction" minerals that show inclined extinction, and for some sections of nominally "inclined extinction" minerals (for example, if (010) is vertical) that show parallel extinction.
One of my favorite examples of what seems to look like anomalous optical behavior is illustrated in this thin section photo I've attached. Although not illustrative of extinction, it still demonstrates that some observations may not necessarily be as they appear. The large green grain looks like an amphibole, with apparent 60°/120° cleavage angles. But it is fact not an amphibole, but rather aegirine, cut obliquely to the two nominally 90° cleavages (so as in structural geology, it's an "apparent dip" issue). Evaluating only this one grain might mislead a petrographer, as a survey of other grains of this same mineral essentially all show the expected 90° cleavage angles. And amphibole is actually present in the sample, but it's the small bluish-green grains in the matrix.
So the expected rules are important to learn. But the addendum to that is that there are exceptions (whether real or apparent) to every rule... :-)
Sorry, only registered users may post in this forum.
Mindat.org is an outreach project of the Hudson Institute of Mineralogy, a 501(c)(3) not-for-profit organization. Public Relations by Blytheweigh.
Copyright © mindat.org and the Hudson Institute of Mineralogy 1993-2019, except where stated. Mindat.org relies on the contributions of thousands of members and supporters.
Copyright © mindat.org and the Hudson Institute of Mineralogy 1993-2019, except where stated. Mindat.org relies on the contributions of thousands of members and supporters.