Monday, July 13, 2015


The other day I was reading a copy of the Mineralogical Record, a mineral magazine (actually more like a professional journal) that is issued bimonthly and is the most authoritative and widely respected mineral collector's journal in the world; no serious advanced collector would be without it (  I did notice that Wendell Wilson, the Editor-in-Chief and Publisher since ~1977, continues to publish an outstanding magazine with beautiful color photos. 
Vug containing druzy wendwilsonite;  collected from Bou Azer, Ouarzazate Province, Morocco. Specimen acquired December 2015.
My mind often goes in interesting directions and I decided to see if a mineral was named after the Editor.  Sure enough, wendwilsonite was named in 1987 for a hydrated calcium magnesium arsenate (contains the AsO4 radical)—Ca2Mg(AsO4)2-2H2O.  In reading descriptions of the mineral I noted it was often referred to as the “magnesium analogue of roselite.” That perked my mind up since I purchased a small specimen of roselite at an estate sale about three years ago and it was filed under “Africa” in a home drawer. 

Photomicrograph of a crimson red aggregate of roselite crystals.  Width FOV ~ 1 cm.
I originally purchased the specimen since the crystals were nicely formed and were gorgeous rose-red to rose-pink in color.  At that time I guessed the color might be due to cobalt but was not certain.  It turns out that roselite is the cobalt analogue of wendwilsonite; cobalt replaces some of the magnesium [Ca2(Co,Mg)(AsO4)2-2H2O] and thus forms a solid solution series.  Any specimen rich in magnesium is wendwilsonite while the cobalt-rich specimen is roselite. But, wendwilsonite usually contains enough cobalt to give a pink color to the crystals while roselite always has some magnesium and many of the crystals are zoned, and are of a lighter shade of color with a decreasing cobalt content.

Photomicrograph roselite crystals showing color gradation from pink to dark crimson red.  Width FOV 6 mm.
Early on I also assumed that roselite was given its name due to the rose red color.  Wrong!  The mineral was named for Gustav Rose, a German mineralogist.  But could there be some relationship between the color and the name?  Only the authors know that answer.

Roselite crystals are actually quite beautiful with their color and vitreous luster.  They are usually transparent grading to translucent and are fairly soft at ~3.5 (Mohs) with a red streak.  The individual crystals range from tabular to prismatic, often twinned, and seem to occur in aggregates rather than individual and well defined crystals.

I find it interesting that roselite (Monoclinic Crystal System) has the same chemical formula (are therefore dimorphs) as another mineral termed beta-roselite (or β-roselite) that belongs to the Triclinic Mineral System.  They have the same color and everything and I seriously doubt if I could tell the difference; however, β-roselite is a very rare mineral.  The same goes for identifying a red to pink specimen of wendwilsonite!

My specimen of roselite (I have none of wendwilsonite or β-roselite) came from its most famous collecting locality, the mines of the Bou Azzer District, Morocco.  The District contains over 60 ore bodies enriched with cobalt and nickel.  Mining of cobalt, nickel, arsenates, gold and silver started in 1928 and in 2006 the mines produced about 8 percent of the annual world cobalt production (Hawkins, 2006).  The ore bodies are associated serpentinites of a Precambrian ophiolite sequence in contact with igneous intrusions and volcanic rocks (Ahmed and others, 2009).  In other words magmatic fluids from the intrusions interacted with upper mantle peridotites (olivine- and pyroxene- rich) and rocks called ophiolites that are pieces of the earth’s upper mantle and ocean floor.  This sequence of rocks seems to date back to the late Precambrian (700 to 600 Ma) when the plates building what is now Africa were banging into each other.  Geologists know this event as the Pan-African Orogeny.  In addition, these rocks were later uplifted by compressional mountain building during the late Paleozoic (Hercynian Orogeny) (Ahmed and others, 2009).  So, the enrichment of primary sulfides came from hydrothermal interaction of magmatic solutions with the serpentinites.

Photomicrograph of two nicely flattened erythrite crystal, common for the mineral.  Crystal<--------> is perpendicular to camera and specimen surface while second noted crystal <--------* is parallel to specimen surface.  Length of longest crystal ~5 mm.

Photomicrograph clusters of erythrite sheaths and radial aggregates from Bou Azzer..  Note different crystal form than roselite above.

Bou Azzer also produces another beautiful cobalt-colored mineral: erythrite, a hydrated cobalt arsenate [Co3(AsO4)2-8H2O].  It also has that crimson to red to pink-violet color, is translucent to transparent, and is very soft at ~2 (Mohs).  Crystals, usually striated and prismatic, are flattened and these factors, plus the softness, distinguish erythrite from roselite.  However, they are found together at Bou Azzer along with a variety of other cobalt-rich minerals (cobaltite) and varieties (cobaltoan calcite; cobaltoan dolomite). Since erythrite and roselite are secondary minerals in the oxide zone they most likely oxidized from one of the primary sulfides such as cobaltite [(Co,Fe)AsS]. 

There is also a second variety of erythrite in the mineral world commonly known as cobalt bloom where the specimens contain cobalt as a druzy-like coating of earthy, non-crystalline material.  It still has the pink to red color but is much “duller” that the nice crystals since it is sort of a weathering crust.

Pink cobalt bloom (erythrite) on specimen from Cobalt, Ontario, Canada.  Width FOV ~7 mm.
Since Bou Azzer also has abundant nickel it would be worthwhile to note that erythrite is in a complete solid solution series with annabergite---the nickel ion substitutes for cobalt ions: Ni3(AsO4)2-8(H20). Evidently there are less attractive “middle members” of the series that I have not observed.  Annabergite also occurs in two forms: 1) the common druzy, dull and earthy weathering crust called nickel bloom; and 2) very nice apple green, translucent to transparent, vitreous crystals.  The crystals are either the flattened blades like erythrite, or acicular masses, often in radial aggregates. My specimen of annabergite does not come from Bou Azzer but from the Lavrion District in Greece and contains both cobalt bloom and a mass of tiny acicular crystals.

Green nickel bloom (NB; annabergite) and massive tiny, apple-green crystals of annabergite (A).  Width FOV ~ 5 mm.
Erythrite is also the cobalt analogue (isostructural; same structure but different chemistry), and annabergite is the nickel analogue, of the rare hydrated arsenates kottigite (zinc), parasymplesite (Fe), and hornsite (Mg).
I have a third specimen of erythrite collected from near “Cobalt, Canada,” one of those old prolific mining areas in Ontario.  Regardless of its name, silver was the major metallic commodity with production starting in the early 1900s and ceasing in the 1930s.  At one time the Cobalt District mines were the world’s largest producer of silver and total production over the years totaled nearly a thousand tons.  The silver was associated with nickel and arsenic minerals like skutterudite and very little of these toxic elements were removed from the landscape; there are a host of environmental problems today.  

I was unable to locate exact production figures of cobalt and nickel for the Cobalt District; however, Young and Perrone (2013) noted that “small high-grade deposits of nickel-cobalt arsenides furnish significant quantities of cobalt. Arsenide ores from Cobalt, Ontario, gave Canada world leadership in production for the period 1905-25. Cobalt output from this area stopped in 1971 but was reactivated in 1995 as a primary production…Canada produced 2013 t of cobalt as recoverable metal…in 2000.”

Specimen collected from Cobalt, Ontario, Canada, with pink cobalt bloom (C), skutterudite (S), arsenic-deficient skutterudite (Sm “smaltite”), and ?annabergite (A). Width FOV ~2.2 cm.  
My specimen has small areas of cobalt bloom along with several “globs” of skutterudite (cobalt iron nickel arsenide; see Post April 21, 2013) and an old label noting the presence of “smaltite an arsenic-deficient skutterudite.”  At once time smaltite was recognized as a distinct mineral that was lighter in color, more tin white in color, than skutterudite.  However, smaltite is now considered as a variety of skutterudite. 

Photomicrograph green to clear lathes of ?annabergite from specimen noted above. Width FOV ~ 9 mm.
The Canadian specimen also exhibits several slender lathes and sprays that have a green to light green to almost clear color.  My wild guess is that they are lathes of annabergite!

Ahmed, A.H., S. Arai, and M. Ikenne, 2009, Mineralogy and Paragenesis of the Co-Ni Arsenide Ores of Bou Azzer,Anti-Atlas, Morocco: Economic Geology, v. 104.

Hawkins, M., 2006, Cobalt News: The Cobalt Development Institute,

Young, R.S. and L. Perrone, 2013, Cobalt: