American Mineralogist, Volume 67, pages 156-169, 1982

New data on and discreditationof "texasite," "albrittonite," "cuproartiniter" "cuprohydromagnesite,"and "yttromicrolite," with correcteddata on nickelbischofite, rowlandite,and yttrocrasite

DoNelo R. Peecon Department of GeologicalSciences The Universily of Michigan Ann Arbor, Michigan 48109

Wrlu.r.rrr B. SrnauoNs,Jn. Department of Earth Sciences Universityof New Orleans New Orleans,Louisiana 70122

Eruc J. EssBNeexo E. Wu. HnrNnrcH Department of GeologicalSciences The University of Michigan Ann Arbor, Michigan 48109

Abstract

Dataare presented which show that "texasite" and "albrittonite" exist only as synthetic phasesand that "cuproartinite"and "cuprohydromagnesite" are not knownto existin any form. "Yttromicrolite" is a heterogeneousmixture of yttrianmicrolite and tantalite. These five phaseshave therefore been discredited as minerals.Major parts of the descriptionsof the mineralsrowlandite, yttrocrasite, and nickelbischofitefrom Llano County, Texas shouldalso be revised. We conclude, among other things, that the formula of rowlanditeis stillin question,that yttrocrasite has not yet been definitively shown to bea memberof the euxenitegroup and that nickelbischofitedoes not occur at the OxfordQuarry, Llano County,Texas. It is suggestedthat procedures for accreditationof a newmineral should includedeposition of a type or cotypespecimen in matrix to a suitableinstitution in sufficientquantity for X-ray diffractionor chemicalanalysis, coupled with the complete manuscriptto be publishedat a later time.

fntroduction Commissionon New Minerals and Mineral Names (Kato, written communication,1981). We have critically re-examined "texasite" (Crook, 1977a), "albrittonite" (Crook and Mar- Data on specificminerals cotty, 1978),nickelbischofite (Crook and Jambor, 1979),rowlandite (Crook et a\.,1978), yttrocrasite "Texasite" (Crook, 1977b), "yttromicrolite" (Crook, 1979), Crook (1977a)states that "texasite" (definedas "cuprohydromagnesite"and "cuproartinite" (Os- PrzOzSO+)is a "supergene alteration product of waldandCrook, 1979).Wefindthataspectsofallof primary rare-earthminerals" occurring in pegma- thesepapers are subjectto question.The resultsof tites of the Llano region of Texas. The primary our studies as presentedas part of this paper led us parent minerals contain typical undifferentiated to recommendthat the minerals"texasite", "albrit- suites of rare-earth elements. That "texasite" tonite", "cuproartinite", "cuprohydromagnesite", shouldhave such an origin, and that Pr shouldbe so and "yttromicrolite" should be discredited. This highly concentratedas to represent>99.9% of the recommendationhas been approvedby the I.M.A. rare earth elementspresent is geochemicallyunrea- 0003-0M>v82l0r02-0156$02.00 156 157

sonable.We believe that the explanation for this is 10,000/1so that a changeof5 in the ratio represents simply that synthetic Pr2O2SOahas been described a change of only 100-200 ppm in the minor rare as though it were a natural material. earth element, differences to be expected in a Dr. John Haschke, a former faculty member of synthetic material rich in Pr. As the synthetic the Department of Chemistry, University of Michi- sample has been purified commercially (Research gan, synthesizeda number of rare-earth compounds Chemicals, Phoenix, Arizona) using advanced ion and donated samples to Peacor prior to Crook's exchangeprocedures to obtain high purity praseo- time of residence at The University of Michigan. dymium, it is unreasonablethat a natural process Among these was a container of crystals of Pr2 could achieve higher levels of purity. These data O2SO4(identical in appearanceto crystals of "texa- therefore demonstrate that the compositions of site") which served as the source of material for X- "texasite" and the synthetic Pr2O2SO4(which was ray crystallographic studies (Peacor, unpubl. data). synthesizedin the samelaboratory as the Pr2O2SOa This containerofcrystals could not be found subse- which is missing from Peacor's laboratory) show a quent to the publication of the description of "texa- close similarity. These data also verify that the site" by Crook, although other samples of material claimed by Crook to be naturally occurring Haschke's synthetic rare-earth compounds were "texasite" is indeed nearly pure Pr2O2SO+.The still present in Peacor's laboratory. This demon- incredibly high purity reflected in his analysescan- strates that crystals of this extremely rare com- not simply be ascribed to analytical error on what pound were available in the laboratory where the might actually be impure natural material. "texasite" was studied. This is at first sight a Second, the fact that Nd was not detected in remarkable coincidence, but it is compatible with either sample(and thus is present only at levels less our other direct evidence showing the equivalence than 200 ppm, see Table l) is especially notewor- of synthetic and natural materials. thy. Haschke notes (written communication, 1979) We submittedsamples of "texasite" (so identi- that "the probability of finding a texasite sample fied originally as such by Crook) for neutron activa- tion analysis, along with samplesof synthetic Prz- Table l. Data for neutron activation analyses of synthetic (Haschke's O2SO4obtained from John Haschke and said by Pr2O2SOa sample)and "texasite" (Crook's sample) him (written communication, 1979) to have been Rareearth activity ratios synthesizedwith the same starting source materials and as part of the same experiments as the Pr2O2 SOaoriginally given to Peacor. The analyseswere Synthetic "texasite" carried out by Mr. John Jones of the Michigan Memorial Phoenix Project, The University of Mich- PrlTb 2.12 2.06 Pr/Dy 4.1 't3.25.2 igan.Results are presentedin part in Table l. These PrlSm 16.8 results show that there are striking similarities between Crook's "texasite" and synthetic Pr2 Pr/Ce 5.4 9.4 Pr/Eu 8.8 14.2't4.0 O2SO4.In the following discussionwe emphasize: PrlLa 14.3 (1) generalsimilarities, (2) Nd/Pr ratios, and (3) Au and Ag values. EulDy 0.06 0.04 Sm/Dy 0.24 0.39 First, there are clear, general similarities between CelDy 0.76 0.55 the analyses of Crook's type "texasite" and Haschke's synthetic phase. DylTb 1.24 0.97 Where rare earth ele- Sm/Tb 0.30 0.38 ments are relatively high in one they are relatively EulTb 0.07 0.04 high in the other. There are some exceptions to this rule, but since the analyseswere obtained on excep- Estimatedabsolute concentrations (ppm)' tionally small samples, the differences are not un- reasonable.Where the differences are highest (for Au 222 l0 Ag not detected not detected PrlEu and Pr/Ce) "texasite" has a lower concentra- (<4) (.4) tion of a given element relative to praseodymium than the sampleobtained from Haschke. It is impor- Tb 450 450 Nd not detected not detected tant to recognize that the absolute values of ratios (<2oo) (<2oo) such as Pr/Tb or Pr/Eu are on the order of 1.000to 158 with the same Nd/Pr ratio as my synthetic sample ed stoichiometric value (19.53Vo)considering that . . . is so incrediblethat it must be excluded.The galenawas used as a standard for S, and consider- 'didymium' separation of (mixture of Nd and Pr) ing the accuracy of normal probe data and correc- into the rather pure elements (neodymium and tion proceduresat The University of Michigan. The praseodymium)is one of the interesting historical quoted weight percents for La2O3, Ce2O3, and eventsof lanthanidechemistry." This difficult sepa- Nd2O3(all less than 0.l0Vo) are below minimum ration is accomplishedusing ion exchangeresins. detectable limits, especially considering X-ray in- Repeatedexchange of progressively more diferen- terferences.This is further indicated by the fact that tiated fractions is necessaryin order to ultimately Nd was not detected in "texasite" by neutron achieve high resolution. Haschke therefore con- activation analysis, as emphasizedabove. In short, cludes that "attainment of the same resolution of these data show that Crook's reported analytical Nd and Pr in a natural system and in a chemical results are almost certainly erroneous. processingfacility is beyond belief." We further Dr. Haschke states (written communication, emphasizethat not only are Nd levels the same in 1979)that: "The lattice parametersand d spacings, both samples,but Nd is not detectedin the $texas- reported on page 1007 of American Mineralogist, ite" sample. Such differentiation of Nd relative to Vol. 62 are my data for Pr2O2SOa.However the I/16 Pr is not credible in a natural mineral. values are not." Haschke further emphasizesthat: Third, of even greater interest is the fact that gold "I have never given Mr. Crook permission to was detectedin both the "texasite" and synthetic publish any of my data or to use any statements Pr2O2SOasamples (Table 1). Suchlevels of gold in a attributed to me." Dr. Haschke also assertsthat rare-earth differentiated mineral are un- other statementsattributed to him by Crook are not reasonable.Haschke synthesized his Pr2O2SOaun- correct. For example Haschke says that (written der hydrothermalconditions in gold capsulescon- communication.1979. in contrastto Crook's state- taining HNO3 at 15,000psi pressure and 500'C menton page1008 of AmericanMineralogist): ". . . (written communication. 1979). Such conditions I do not (and never have) attributed the existenceof could be expectedto yield low but variable concen- a pure praseodymium mineral to crystalline stabil- trations of gold as contaminantsin the run products. ity. The structuresof Nd2O2SOaandPr2O2SOa are Thesedata are similar in every respectto thosefor identical, and in my opinion, segregationof Pr could the syntheticPr2O2SOa present in Peacor'slabora- never be attributed to a unique structural stability of tory and synthesizedby Haschke.We alsonote that the praseodymiumcompound." although the gold contents are different, the magni- In addition to the report of "texasite" from the tude of the difference is not unusual for a contami- Clear Creek and Rode Ranch pegmatitesin Texas, nant in specimenssynthesized under the conditions Crook (1978a) also reported it from the Luster notedby Haschke.It is the presenceof ppm levels pegmatitein Colorado.The Colorado "texasite" is of gold in the so-called natural phase which is not reported to have somewhathigher levels of La, Ce, credible. In addition, Jones (written communica- and Nd, but otherwise is rather similar to the Texas tion, 1979)has analyzed(using NAA) a number of occurrence.While we have not been able to analyze pegmatite samplesfrom the Llano region and, in- the Colorado "texasite" for rare earth and other deed,no gold was detectable. elements,most of the objectionsraised above for The chemical analyseson "texasite" given by the type "texasite" also apply to this occurrence. Crook (1977a) show average values of: PrzO:, The discreditation of "texasite" has been ap- 8O.6lVo;La2O3, 0.08%; Ce2O3, 0.03Vo; Nd2O3, proved by the I.M.A. Commissionon New Miner- 0.02%; SO3, 19.52Vo.Crook states that he used als and Mineral Names (Kato, written communica- syntheticPrzOzSO+ as a standardfor his Pr analysis tion, 1981). but galenafor S. The use ofgalena as a standardfor S in "texasite" insteadof Pr2O2SOais surprisingin "Albrittonite" that the synthetic sulfate of presumed identical "Albrittonite" was describedby Crook and Mar- compositionto "texasite" was used as a Pr stan- cotty (1978).1We have found discrepanciesin the dard. Galenais a poor standardfor S and PrzOzSO+ is an excellent one in this case considering that the rMarcotty (written communication, 1978)has informed us that averageatomic numbers of the two compounds are her part in the researchwas to obtain powder X-ray patternsand very different. Furthermore, Crook's averagevalue this did not extend to their interpretation. Our comments are for SO3(J9.52%) is remarkablyclose to the expect- addressedtowards the researchattributable to Crook. crystal drawings and related data, in the inherent another drawing entitled "typical morphological unlikelihood of the reported occurrence given the development of albrittonite." This drawing is en- high solubility in H2O and the miscibility with tirely different than those pictured as Figure 1 It is nickelbischofite,and in inconsistenciesin the analy- clear that the morphology of "albrittonite" crystals sis. has been misrepresented. We reproduce Figure I of the "albrittonite" J. L. Jambor(written communication,1980) pro- (CoClz ' 6H2O) description as part of our Figure 1. vided us with a copy of a letter from Crook to Beside it we have placed direct reproductions of Jambor dated July 5, ln8. In this letter Crook certain figures from Vol. Il of The System of states: "Now, my diagram of albrittonite. I agree ,Seventh Edition (Palacheet al.,l95l). that the drawing doebnot look like it has a (001)face These two figures represent crystals of liroconite but the measureddata are: of The Systemof Mineralogy) and paralaur- @.921 A ionite (p. 65). Detailed comparison shows that the 0 9 6z Pz:B 90000' 27"12', 62"48', 90000' 62"47', "albrittonite" crystal drawings and those of para- laurionite and liroconite are identical, except for the This fits the indices for (001) well-perhaps the style of lettering on the faces. We therefore con- drawing is just a poor projection." The diagram to clude that the "albrittonite" crystal drawings have which Crook is referring is the one which is identi- been copied from The System of Mineralogy. Fur- cal to the one shown with the description of the thermore, an early version of the albrittonite paper mineral paralaurionite in The System of Mineral- on record with the Smithsonian Institution with ogy, Volume II (1951).On page64 of Volume II of albrittonite specimen accession papers. contains The System of Mineralogy (1951)also under the

LIROCONITE ( FromThe System of Minerology,Vol.II, p.921)

Cornwall.

(e) o 'l-

Fig l. Typical morphologicaldevelopment of albrittonite.A = @) common habit. B = occasionallydeveloped [orm. Forms areo = {1001.c = l00ll.e = {0lll, nr = {ll0}

Laurium.Twinned on I100|'

PARALAURIONITE (From The Systemof Minerology,Vol.ll, p.65) Fig.l. Copyof Fig.I fromCrook and Marcotty (1978), said to be"typical morphological development of albrittonite" (on the left), togetherwith copiesof figures(on the right)of crystalsof liroconiteand paralaurionite from Vol. II of TheSystem of Mineralogy. PEACOR ET AL.: NEW DATA ON "TEXASITE." ETC. description of paralaurionite, is a table of data Here again the correspondence between Crook's dealingwith the morphology of paralaurionite crys- interfacial anglesand those reported inThe System tals. The first line reads: of Mineralogy for another unrelated mineral is clear.On the other hand,Crook's valuesas givenin ' Op+2 with "albrittonite" are incompati- c 001 his notes dealing 90000'27"12112' 62'47ll2' on "albrittonite." It C A ble with his own unit cell data h:B therefore appears as if interaxial angles for liroc- 90000' 62"47112' onite and paralaurionite have been used, changed These angles are essentially identical with those slightly, and reported as values for "albrittonite." given above, angles which Crook states he mea- "Albrittonite" is approximately twice as soluble sured for "albrittonite." This indicates that the in water as, NaCi (and nickelbischofite, which interfacial anglesfor paralauiionite have been sub- Crook and Marcotty (1978) state to be associated stituted for those of "albrittonite" and castsfurthe with "albrittonite." is seven times as soluble ac- doubt upon the crystal drawings of "albrittonite. cording to Weast, 1980).These minerals are said to Crook (written communication, 1980) has pre- occur due to the "influx of water through ground pared hand-drawn crystal drawings which are simi- water, rain, etc. causes albrittonite to dissolve lar to those which he figured in his paper on readily and to be reprecipitated upon the resump- "albrittonite," and which he states are his notes tion of arid conditions." Dr. R. L. Freed (written relating to measurementsof "albrittonite" crystals. communication, 1980)of Trinity University, San On his drawings are morphologic data which he Antonio, Texas, questionsthis statementand states states he measured from "albrittonite" crystals that "The Llano region is really not all that arid, using an optical goniometer. Using the values on receiving approximately 28 inches of rain per Crook's drawing, which resemblethe Systernfigure year." The extreme solubility of CoCl2 ' 6HzO of liroconite, we calculate the following ratios: makes the Llano occulrence unlikely although not b/(a sin 9) : 0'58 impossible.However, Dr. Freed notes that he has visited the serpentine quarry reported to be the and sourceof "albrittonite" (and nickelbischofite)in an (c sin 9/b) : 1'31 attempt to collect it and "to search for the vein of cobaltite, linnaeite, siegenite, and nipkeline men- For comparison, we calculate from the published tioned in Crook and Marcotty (1978).The Llano crystallographic data on "albrittonite" and from region is unusual in its lack of sulfide mineraliza- equivalent values for liroconite that: tion; the report of primary sulfide and arsenide minerals is noteworthy in its own right." The latter ,n"llit!fi"ll"r,". "albrittonite" Loluie II (1951) Crook and l'4arcotty (1978) minerals were reported by Crook and Marcotty (1978)to serve as the primary sourceof the second- b/(a sinB) 0.60 0.80 ary "albrittonite" and nickelbischofite. Freed (writ- (c sinB)/b l.3l 0.93 ten communication, 1980)notes that "The minerals This shows that the measurements(Crook's, writ- and the vein were reported from the north side of ten communication, 1980) could not possibly be the quarry and I spent approximately three hours obtainedfor "albrittonite." However Crook's in- searching'thisnorth side. Two other scientists,a terfacial angles predict axial ratios in agreement mineralogistand geologist,were with me during the with those of liroconite. search,and they also were unable to find either the A seconddrawing by Crook (written communica- primary vein or the new minerals." Freed also tion, 1980)resembles the figure in The System of notes that "The Oxford serpentinequatry has been Mineralogy, Volume II (1951) of a paralaurionite a field trip stop for studentsand professionalgeolo- crystal. It also has numbers measured by Crook gists for many years. It is amazing that minerals of from an "albrittonite" crystal. Comparisonssimilar such distinctive colors have not been reported to those given above are: either prior to 1976, or after 1976." Crook and Marcotty (1978)reported the occurrenceof zaratite, Calculated from Calculated fron Calculated frm data on oala on erythrite, and annabergitewith "albrittonite" and valueson paralaurionite "al bri ttonite" Crook's drawinq The sys. of l'lin. nickelbischofite and it is to those colorful minerals b/(a sin B) 0.40 0.41 0.80 that Freed refers. They are not readily water-

B 117.07" 117.?2" 97.25" soluble and once formed would not be easilv dis- l6l solved.These relations cast doubt on the validity of sample regardless of the beam width employed in the natural occuffence of "albrittonite" (CoCl2 analysis. Accurate quantitative microprobe analy- 6H2O) and nickelbischofite(NiCl2 ' 6H2O) at the sis is simply not possibleunder theseconditions. Oxford Quarry, Llano County, Texas.We note that Another seriousanalytical problem concernsthe a quantity of material provided to his co-author, analysesof "albrittonite" for H2O. The two report- Marcotty (personalcommunication), is identical in ed values(45.61 and 45.16percent) are surprisingly appearanceto syntheticCoCl2 . 6H2Oand, consid- closeto the ideal value of 45.43percent. Crook and ering their statementsregarding its origin, remark- Marcotty (1978)state that "water was measuredby ably free of any inclusions or contaminants.We measuringtotal weight loss on ignition minus the conclude that the above evidence suggeststhat the chlorinecontent obtained from the probe analysis" "albrittonite" and nickelbischofitestated to be nat- (Crook and Marcotty, 1978).This is unlikely ac- urally occurring at Oxford, Texas are instead syn- cording to M. H. Hey, who questions (written thetic material. communication,1980) such a statementof proce- BecauseCoCl2 . 6H2O("albrittonite") and NiCl2 dure in the strongestterms. Hey states that he has ' 6H2O(nickelbischofite) are isostructuralthey mhy carried out the experiment and finds that ignition be expectedto exhibit mutual solid solution.Crook below approximately 900"C gives a residue of and Marcotty (1978)and Crook and Jambor (1979) Co3Oawhich correspondsto a total weight loss of report that they occur together and were derived 66.27Vo,or to a weight loss on ignition minusthe Cl from the same vein of primary sulfides at the content of 36.5Vo.At approximately900"C Co3Oa Oxford Quarry. Becausehis analysesof "albritton- loses oxygen and is converted to CoO. Ignition ite" averageonly 0.3VoNi (and nickelbischofiteis abovethis temperatureresults in a total weight loss reportedto have only 0.57oCo) it appearedthat our of 68.5Vo,which minus chlorine, yields a value of prediction of solid solution was at variance with 38.7Vo.Neither result correspondsto the weight Crook's analytical data. We therefore prepared percentagesof water (45.61and 45.16).Hey notes solutions of approximately equal amounts of both that only if the ignition product is pure cobalt metal compoundsand allowed the water to slowly evapo- would their value for HzO be correct, "and to get a rate so that crystalswere formed. Well-developed, residueof Co metal it would be necessaryfor the single-phasecrystals grew. They were shown to be ignition to take place in a reducing atmosphere- of an intermediate composition by X-ray diffrac- H2, or possiblyCO-and then the metal would very tion. This demonstratesthat extensivemutual solid likely be pyrophoric." We therefore conclude that solutionexists betweenCoCl2 . 6H2O and NiCl2 . the analytical data for H2O reported for "albritton- 6H2O. The coexistenceof nearly pure "albritton- ite" are invalid. ite" and nickelbischofiteunder the conditions re- Hey has also pointed out (written communica- ported is highly unlikely. tion) that there appears to be a problem with the The analytical data for "albrittonite" (and nickel- optical data for "albrittonite" and he states that: - bischofite)are themselvessubject to question.Be- "^y- B 0.026,B c 0.025is not consistentwith 2V cause they are of remarkably high accuracy, we 53o" (the valuesgiven by Crook and Marcotty.) have considerabledifficulty understandinghow this In summary we note that the reported crystallo- could have been accomplished.First, "albritton- graphic data, geologic occurrence, and chemical ite" is so soluble that during the polishing processit analysesfor "albrittonite" are questionable,and do partially dissolveseven in alcohol and kerosene.No not exist as presented.It is for these reasonsthat mention of this difficulty was noted. Marcotty (writ- "albrittonite" has been discreditedas a valid miner- ten communication,1980) noted a similar difficulty al species, in preparing smears of CoCl2 6H2O for X-ray diffraction studies.We obtaineda mediocrepolish Nickelbischofite with great difficulty only after dry grinding and Nickelbischofite was described by Crook and polishing. More significant, however, is the fact Jambor(1979).2In the precedingsection we raised that in our attempts to duplicate Crook's micro- seriousquestions regarding "albrittonite" data,and probe analyses,these sampleswere found to be at the sametime noted that many of those questions highly unstableunder the electron beam. Under the zJambor(written communication, 1980)reports that he only conditions stated by them (12 kV, 150 na), we worked on the Amos, Quebecnickelbischofite and did not obtain observedit to "boil" and exhibit signsof extensive any data on Crook's samplesfrom the Oxford Quarry, Llano beamdamage the instant the beam is exposedto the County, Texas. were equally applicable to nickelbischofite. We conclusion of the paper is that rowlandite has the shalltherefore not repeatthem here. However, we formula RE3(SiO4)2(F,OH).We are concernedhere emphasizethat those questionsrelate only to the only with the chemical data which was reported by occurrence of nickelbischofite at the Oxford ser- Crook et al. fi97$. pentinequarry, Llano County, Texas, and to data Two chemicalanalyses of rowlanditeare given in reported by Crook and Jambor (1979)from Texas Table2 of Crook et al. (197$. Theseare similarand material. Independently reported occurrencesof sumto 99.11and 98.51weight percent, respective- nickelbischofiteare those of Shima (1957), who ly. We have normalizedanalysis I to 2.215 Si (as detecteda compoundwhich may be NiCl2 . 6H2Oin givenin the "calculatedformula") and find reason- volcanicsulfate-rich sublimates, and Jambor(1975) ableagreement for most elementswith valuesin the who noted the mineral on drill core from the Du- "calculatedformula" exceptfor fluorine.Our value mont ultramaficintrusion, Amos, Quebec.Jambor for molar fluorine is 1.85 times as large as that (1975)obtained an excellentX-ray powder pattern given, and shows the latter to be in seriouserror, for the Amos mineral and noted at that time that it based on the reported weight percent of fluorine. correspondedto NiCl2 . 6HzO. Subsequentconfir- Whereasvalues for F. Cl. and OH are summedto mation of Ni and Cl in the Amos mineral, as 0.997in the paper, they sum to 1.556according to reported in Crook and Jambor (1979),was obtained our calculations,thus suggestingserious error ei- at the GeologicalSurvey of Canada in 1977, and ther in the analyses,orin the formula obtainedfrom communicationsconcerning the naming of the com- it. If the formula is correct. on the other hand. the pound had been initiated by Jambor with the Chair- summationfor the analysis (given the necessary man, Commissionon New Minerals and Mineral changein fluorine and oxygen) would be unsatisfac- Names,I.M.A., well prior to Crook's independent tory. We note that the samekinds of errors involv- submissionof data for NiCl2 . 6HzO.The statusof ing fluorine were made by Crook (1977c)for analy- nickelbischofite as a naturally occurring mineral ses of rare-earthfluorides in his Master's Thesis therefore should be retained. Occurrenceof a high- (Crook, 1977c).This implies a pattern of errors in ly soluble mineral such as nickelbischofiteis not which the formula content of fluorine is generally surprisingin drill cores or in volcanic sublimates, that expectedfor the mineral, yet the elemental unlike the surface exposuresat the Oxford quarry. weight percent from which the formula was suppos- Soluble minerals are less likely to be dissolved edly derived is not in agreementwith the formula inside a solid rock sample.The rapid supersatura- given. This is the kind of error expected during tion attendantupon volcanicdegassing allows tran- hand calculation,for instance,of a weight percent sient formation of highly soluble and perhapsunsta- analysisfrom a previously assumedformula, but ble minerals.We believethat only the occurrencein not the kind of error obtained using the e,upeon VII Llano County, Texas and data from such material program which produces a correct molar formula are invalid. Physicalproperties and X-ray crystal- for F from the element weight percent as obtained lography of nickelbischofite as given in Crook and from a microprobe analysis. Jambor (1979) conform with data for synthetic The electron microprobe standardsfor rare-earth NiCl2 . 6H2O as given by Groth (1906),Kleinberg elementsCrook has usedare subjectto question.A (1969),Swanson et al. (1974),and Winchell and statementis made(Crook et a1.,1978)that: "Ana- Winchell (1964)except as follows (Jambor, written lyzed specimensof gadolinite and yttrialite (Rode communication,1980): morphological forms, stated Ranch pegmatite, Texas) and synthetic rare-earth as{100}, {100}, {011} should read {100}, {110}, {001}, saltswere usedas electronmicroprobe standards." {20I}; natural occurrenceof thesecrystals up to 15 No mentionof the correspondencebetween specific mm long is improbable; pleochroism of ct pale elementsand specificstandards is made, as would ^y green, greenshould be a greenishyellow, ^ygreen; be proper. Moreover, Crook (written communica- absorptionT ) c, not cr ) 1. tion, 1979)has indicated that gadolinite was used Rowlandite for the majority of the rare-earthelements. Because gadolinite is subject to metamictization it generally We refer here to the data relating to rowlandite in shouldbe a poor probe standard.Indeed, a probe a papergiven (1978).3 by Crook et al. The principal standardof the Rode Ranch gadolinite as prepared 3Ewing (personal communication, 1980) reports that Crook by Crook and donatedto us by him was examined obtained all the chemical data for rowlandite. by us on the electron microprobe. It proved to be 163 extremely heterogeneousin the rare-earth elements Table 2. Normalization of yttrocrasite analysis as reported by (1977b) and entirelyunsuitable as a probe standard.Results Crook obtained with it would therefore be unacceptable. This work* Crook Furthermore,Crook's sourcesof the gadoliniteand yttrialite standard analyses are not documented. 0.090 0.104 Mn 0.006 0.007 Lastly, neither gadolinite nor yttrialite contain sig- Fe 0.060 0.092 nificant fluorine so we assumed that a rare-earth It 1.838 1.838 l^l 0.023 0.062 salt or was used as a fluorine standard. The Pb 0.005 0.006 natureand sourceofthat standardshould have been T. to 0.001 0.002 listed. In fact identical analyses of rowlandite ap- Nb 0.0004 pearin Crook's thesiswhere he lists only fluorapa- Th 0.102 0.It8 tl 0.025 0.029 tite and fluorite as fluorine standards. Interferences u 1.569 0.942 of rare-earthelement emissionlines Y 0.523 0.601 on the electron microprobe are generally serious. La 0.003 0.003 general LE 0.01I 0.013 Only a brief comment about a method of Pr 0.002 0.002 dealing with such interferences is given by Crook (1977c) (1978), Nd 0.005 0.0007 and Crook et al. and no specific Sm 0.015 0.0l7 details are given despite their significance.Crook Gd 0.034 n nn?o (1977c) (p. 0.001 0.001 notes 32) that the Kcrl line was used for Dy 0.041 0.018 the element yttrium at 15 kV operating voltage. Ho 0.002 0.002 lnterferencesnotwithstanding, there is a serious Er 0.025 0.029 problem in that Y Ka1 radiation has an energy of tm 0.002 0.002 Yb 0.004 0.005 14.956kV. Because a significant overvoltage is Lu 0.003 0.003 requiredto exciteemission lines, Crook's analytical valuesfor Y are subject to question,especially in *All valuesnormalized to 1.838Ti, the value view of the fact that Y is a major element in his reported by Crook. analysesof rowlandite (29.44 and 26.18%oYzOr). We have tuned Y Kct using syntheticY2O3 on the electronmicroprobe at 0.015pa and plotted counts analysis is ascribed to a specimen from Clear vs. operatingkilovoltage. It is clear that Y is not Creek. Furthermore, the analysis of yttrocrasite in efficiently excited below 25 kV and that no counts Crook 0977c\ said to be for material from Clear above backgroundlevels are generatedbelow 18 Creek is duplicatedby Crook (1976).These asser- kV. No yttrium Ka X-rays can be excited at all tions seemto be clear inconsistencies. below 17.037kV (Liebhafsky et al., 1972), yet We have normalized the yttrocrasite analysis Crook has claimed (written communication, 1977, givenin Crook's (1977b)Table I to 1.838Ti atoms, 1978)that he was able to generate48,716 counts the number given as his normalized value. In our above background for yttrium Ko,, on YzOt at 15 Table 2 we compare Crook's values with those kV in an evaluation ofinterferences he provided for which we calculate from the analysis of his paper. the authors. Crook's (1977b,1977c, 1978a, 1978b, Major differences are apparent, as are apparently 1979)and Crook et al.'s (1978)data for yttrium random small differences. could not have been measured as claimed and it Using the lattice parameters stated by Crook appears that they were derived without measure- (1976)to have been obtained by least-squaresre- mentsever actually having been made. finement, we have calculated d-values (Table 3 of this paper). Our Table 3 also contains the d(Obs) Yttrocrasite and d(Calc) values given by Crook in his Table 1. We refer here to the paper on yttrocrasite by There are discrepancies in values of d(Calc) due Crook (1977b).The analysisgiven in Crook's Table simply to an interchangein the lattice parametersb 2 and implied to be of yttrocrasite from Clear and c. Aside from the fact that this is an error which Creek, is duplicatedin Table 18 of Crook (1977c). should be noted for the record, it is difficult to There the analysis is reported to be of yttrocrasite understandhow Crook's erroneous values could from the type locality obtained from specimen have occurred since the lattice parameters are NMNH #R5031, not from Clear Creek; a second basedon a least squaresrefinement. Table 3. Powder diffraction data for yttrocrasite as reported by yttrocrasite is a member of the euxenite group, Crook (1977b),with recalculatedd-values. someor all of the data must be invalid at best, and this d(obs) (Crook) d(calc) (Crook) d(catc) this work must include the X-ray diffraction data. Be- causethe X-ray results are basedon least-squares 'll0200 5.455 6.431 6.431 4.331 4.310 4.505 refinementof single-crystaldata, and becausethe 310 3.121 3.126 3.200 211 2.954 2.944 2.945 parametersso-derived lead to satisfactory indexing 002 2.401 2.403 2.286 of the separate powder diffraction data (with cor- 102 2.369 (.JOJ 2.250 rectionsas notedabove), the data are contradictory 5',10 2.241 2.241 2.268 600 2.147 2.144 2.14? and incornpatible. We therefore question the valid- 420 'rI .865 I .863 1.926 022 .658 t -of,o t.b5/ ity of all of the X-ray diffraction data. The chemical analysis of yttrocrasite given by 'll,1091? 'II .189 1.186 L175 .132 'I 'II.l33 '|.l 36 Crook (1976, 1977b,1977c) is subject to the same 414 I .094 .093 .051 generalproblems (i. e., interferences,inappropriate standards,etc.) as have been discussedin the case of rowl.andite.Again, Crook (1977c)claims analysis A more serious problem with the X-ray diffrac- of yttrium using Y Kol at 15 kV which is impossi- tion data concerns the magnitudes of the unit cell ble. parameters(a : 12.862,b : 4.810, c : 4.5714). We summarize the data for yttrocrasite by noting These were said by Crook (1977b)to have been that there are serious inconsistenciesreported by obtainedby least-squaresrefinement of single-crys- Crosk (l97b) on yttrocrasite, especially concern- tal data, such data serving also to identify the space ing X-ray diffraction and analytical data. This raises group as Pbcn. All of Crook's data (chemical,X- sErisu* questionsregarding the validity of all of the ray, etc.) imply that yttrocrasiteis a memberof the deta h* the paper in question. Whether or not euxenite group. The lattice parameters of typical ythg€fasife is a valid member of the euxenite group members of this group are more than ten percent ' tterefore still rernains to be shown. greater than those given for yttrocrasite. Crook notes that "The shortenedcell parametersin yttro- "Yt ramierolite'' crasite must be due to considerablecorrugation of We,refsrto the paper by Crook (1979)in which he the zig-zagchains of B (Ti) octahedraresulting in an pulFsrts to show that hjelmite should be redefined overall contraction of the strucfure." Given as'?ttrgmtcrolite." We obtained the cotype sam- Crook's lattice parametersand compositionshow- ple of "yttromicrolite" frorn the Smithsonian Insti- ing 24 (O,OH) per unit cell, each (O,OH) ion tutisn. X-ray examination showed that it was large- occupiesan effective volume of 11.8A3(Cell vol- ly ar-x-rorphouswith only two lines near to the umel24).However, we calculate, assumingan anion principal lines of presumably derived from a radius of 1.3A, that in an ideally closest-packed srm$ transparent inclusion on one side of the array of anions an anion effectively occupies sa;lrry,Ie.Elostron microprobe analysis revealedit to 12.4A3.This is 0.5 A3 larger than the value for ,bea mixturerof different phases.At least 95 percent yttrocrasite, implying that the parametersfor yttro- ofthe matcriat esntained only Ca and S as elements crasite are incorrect. In addition, as cations are wfth e + t0 and is presumably a calcium sulfate. substituted into a closest-packedarray, expansion Contained in it were five rnicron blebs of heteroge- generally occurs. In cubic-closest-packed spinel nwBp rdcbotantalates,with variable Ca, IJ, T, Th, (with sites occupied by Mg and Al) the effective *nd, F€. lV'hile difficult to analyze due to their volurneper oxygen atom is 16.6A3for example.In heterogeneity,these appearedto have at least two euxeniteit is approximately 1743. Lastly, distor- iompositions, one similar to niobian tantalite and tions such as the "comrgations" hypothesizedby one similar to a niobian microlite. The microlite Crook result in increasesin effective anion volumes contained variable amounts of uranium, yttrium, in ideal closest-packedarrays. From these and titanium, cerium, neodymium, samarium,and thori- other points of view, it is clear that the unit cell um but alwayshad Ca > Y, U > Th. Crook did not parametersgiven by Crook for yttrocrasite lead to a identify a calcium sulfate, nor did he describe cell volume which is incompatible with the euxen- heterogeneousniobotantalates. His method of ana- ite-group structure and formula. As the other data lyzing yttrium in other paperson yttrian compounds given in the paper in question appear to show that suggeststhat the "yttromicrolite" analysisis alsoin 165

Pruerrolmp

Fig. I Morphologicaldevelopment of cuprohydromagnesite. FormsareA = {010},c = lml},r = {310}.

( FrornThe Systemof MinerologY, Vol.II, p.325 )

Fig. 2. Copy of Fig. I from Oswald and Crook (1979) (on the left) said to be "morphological development of cuprohydromagnesite," together with a copy of a figure of a pinakiolite crystal (on the right) from Vol. II of The System af Mineralogy. error. We concludethat the type "yttromicrolite" reported for "cuprohydromagnesite" and those of is an amorphous mixture of calcium sulfate, tanta- drawing of pinakiolite figured on page 325 of Vol. lite and heterogeneousmicrolite. It has been dis- asbeing orthorhombic with aibic :0.8339:1:0.5881 creditedby the I.M.A. Commissionon New Miner- (Palacheet al., 1951)whereas cuprohydromagne- als and Mineral Names (Kato, written communica- site is reportedto be monoclinicwith a : 10.653,b tion). = 9.141,c : 8.570Aand p : 115"23'(Oswald and Crook, 1979) for which the corresponding axial " Cuprohydromagnesite" and " Cuproartinite" ratiois a:b:c: : 1.165:1:0.938.There is thereforeno We refer here to the paper by Oswald and Crook correspondencebetween axial ratios even though (1979).4Figure I of the paper is reproduced as part the crystal morphologies appear to be identical. of our Figure 2. It was labelled as "Morphological This apparentcopying of a figure of another mineral developmentof cuprohydromagnesite.Forms are b without reference,and apparentinaccurate descrip- : {010},c = i001}, and x : {310}." We have also tion of so-called "cuprohydromagnesite" crystals reproduced as part of our Figure 2 the crystal in part led us to re-examine other data regarding drawing of pinakilolite figured on page 325 of Vol. "cuprohydromagnesite" and "cuproartinite. " Il of The System of Mineralogy (Palache et al., A type specimenwith both minerals was deposit- 1951).On page324the form indicesare givenas b : ed by Crook in the SmithsonianInstitution (NMNH {010}and x : {310}.The similarity in figuresis even #114188).Mr. Pete J. Dunn of the Smithsonian more remarkable given the significant difference in Institution has informed us (personalcommunica- lattice parameters (and axial ratios) between those tion) that in his study of this specimenhe had been able to detect neither of the minerals in question. aOswald (written communication, 1980)has informed us that His studyincluded the useof Gandolfiphotographs' preliminary he originally collected the specimensand obtained He reported that he did detect X-ray peaks charac- qualitative data on them. However, the final reported morpho- logical, X-ray, analytical, and other data were all independently teristic of nakauriite. We have independently been provided by Crook. unableto detect X-ray lines of either "cuprohydro- magnesite" or "cuproartinite" on that specimen has also provided us with copies of Crook's notes which was kindly lent to us by Dunn. (personal communication, 1980) showing that Crook also had donateda specimento the Uni- Crook obtained four separate,complete analysesof versity of Michigan which he labelled "cuproar- cuprohydromagnesiteand three of cuproartinite, all tinite" from the type locality. Our careful powder of apparentlyexceptional accuracy. We question X-ray diffraction of selectedportions of that sample that such small amounts of materials which are has failed to provide data consistent with that exceedinglydifficult to separatecould lead to such reportedfor "cuproartinite." However, the powder accurateanalytical values. This is especiallysignifi- diffraction patterns of this phase are compatible cant in that even Oswald'scarefully chosen "best" with those of nakauriite (Suzuki et al., 1976).In examplesof these minerals turned out to be largely addition, a number of samples of nakauriite and nakauriite,as we explain below. Such samplesof "cuproartinite" from Gabbs, Nevada were kindly nakauriite could not possibly lead to such apparent- given to us by Dr. David Garske who in turn had ly exceptionally accurate analyses of two other obtainedthem from Oswald (Garske.written com- minerals.It is even reported that DTA, TGA, and munication, 1980).The diffraction patterns of the DTG analyses were carried out and the results sampleslabelled nakauriite are consistentwith the appearto be highly accurate(i.e., closeto theoreti- data for nakauriite; those from samples labelled cal values). Given the statementthat very small "cuproartinite" are also consistentwith nakauriite quantities of material are available and that it is and in no way are compatiblewith Crook's pattern subjectto contamination,these statements of meth- for "cuproartinite." As the descriptionsof speci- ods of analysisfor H2O and CO2must be invalid. mens of "cuproartinite" and nakauriite are very Data are presentedin the paper in question which similar, we inferred that "cuproartinite" could, in imply significantsolid solutionbetween Cu and Mg. fact, be identical with nakauriite despite the fact The ion Cu2+ does not generally enter into solid that Oswaldand Crook's (1979)data for "cuproar- solutionwith ions such as Mg'*. This is becauseof tinite" appearsto be definitively different than that the very different coordination requirements of for nakauriite.Oswald, when initially informed of Cu2*, in part related to the well-known Jahn-Teller theseobservations, agreed that this might indeedbe effect in Cu2* compounds.It would, nevertheless, possible(personal communication). At this point in be reasonableto expect ordered substitution of our researchOswald kindly provided us with all of Cu2* for Mg. However the significantsolid solution his notesand researchmaterials, including samples of Cu2* for Mg as describedin the paperin question which he notedwere his best examplesof supposed is difficult to accept as valid. "cuproartinite" and "cuprohydromagnesite." The In view of the above relations we carried out a samplesso obtainedwere reportedby Oswald(wrir detailed investigation of the materials and samples ten communication,1980) to be identicalto material kindly providedby Oswald.These included vials of originally suppliedby him to Crook. carefully selected material remaining from Os- Both "cuproartinite" and "cuprohydromagne- wald's original qualitative research done in 1972 site" are describedby Oswald and Crook (1979)as which had led to his conclusion that this specific occurring as small crystals, sprays, etc. intimately materialwas probably a new mineral. (Indeed,our associatedwith hydromagnesiteand artinite. Os- work shows that it was a new mineral at that time in wald has confirmed (personal communication, that Oswald's researchpreceded publication of a 1980)that theseand other associatedminerals occur description of nakauriite). Our X-ray diffraction in small grains and are thoroughly intergrown. patterns of this material are in no way related to Indeed,the statementis madeof both mineralsthat those of "cuprohydromagnesite" or "cuproarti- the available quantities of both minerals were so nite," but they are very similar to those of nakaur- smallthat standardchemical analyses could not be iite (Suzuki et al., 1976).We obtained Gandolfi performed thus necessitating electron microprobe patterns for two specimensof "cuprohydromagne- analyses.However COz and H2O were said to be site" and two of "cuproartinite," saidby Oswaldto "measured by gravimetric analysis using absorp- be his best,carefully selectedexamples of each.All tion tubes and acid decomposition" and "by total four patterns are nearly identical. They in no way weight loss on ignition minus analyzed CO2 con- resemble the published patterns (Oswald and tent," respectively. The reported H2O and COz Crook, 1979)for "cuprohydromagnesite" and "cu- values appearto be exceptionally accurate. Oswald proartinite," but they are very similar to the pat- 167 terns reported for nakauriite. Table 4 comparesour lattice parametersa = 9.62(l), b : 6.231(6),c : powder data with that for "cuproartinite" and 7.857(8)Aand B : lll.82(7)o, these final lattice "cuprohydromagnesite."These data strongly indi- parametershaving been obtained by least-squares cated that "cuproartinite" and "cuprohydromagne- refinement of powder diffraction data for nakauri- site" do not exist as valid species,and that speci- ite. The latter was obtainedusing a polycrystalline mens purported to be of them are all nakauriite. mount in a Gandolfi camera. The indexed powder The best of the specimensof so-called"cuprohy- pattern is given in Table 4. The indexing is satisfac- dromagnesite" consisted largely of rather platy, tory for this pattern which was obtained using the blue material, thus matching the original description "best" "cuprohydromagnesite"specimen provid- of "cuprohydromagnesite" and appearing to be ed by Oswald.This showsdefinitively that "cupro- different from the so-called "cuprpartinite" and hydromagnesite"specimens are in fact specimens nakauriite, both of which are described as being of nakauriite. The single-crystal data that we ob- fibrous. A very small, light-blue, transparentcrystal tained for the sample labelled "cuprohydromagne- suitable for single-crystal diffraction studies was site" are for nakauriite. Both the unit cell and separatedfrom this material. Precessionand Weis- powder data represent a significant improvement senbergphotographs show that it is monoclinic with over thoseof Suzuki et al. 0976\. There are minor differences between powder Table 4. Powder diffraction data for a "cuprohydromagnesite" patternsof different specimens.This is especially specimen (nakauriite) provided by Oswald (typical of all of true for diffractometer patterns as obtained using "cuprohydromagnesite" and "cuproartinite" samplesprovided samplesspread on glassslides as opposedto Gan- by him) compared'with the published patterns for dolfi powderpatterns, but we assumethat is largely "cuprohydromagnesite" and "cuproartinite." due to preferred orientation effects. We have often

I observeda line having d :784 in diffractometer ila kau r'l i te Cuprohydmagnes i te2 Cuproa rt i n i te2 patterns, as also noted by Suzuki et al. for some 1/ls d(obs) d(calc) hkl d(obs) I/lo d(obs) t/lo nakauriite. This line cannot be indexed using the 100 7,32 7.29 001 9.63 20 8.30 40 i0 5.10 J.lt 110 6.53 25 6.?5 20 unit cell describedabove. However, this line (and 20 4.81 4.84 101 5.91 100 80 10 4,44 4.47 200 4.810 5 4.60 t0 othersdescribed by Suzuki et al., as being charac- 30 3,93 3.93 102 4.570 i0 4.15 l0 teristicof Type II nakauriite)is readily explainedas 5 3.66 3.55 002 4.281 5 3.800 50 40 3.54 3.54 202 4.259 10 correspondingto hydrotalcite or another member of ?0 3.303 3.301 20I 4.130 5 2.900 20 5 3.111 3.116 020 3.870 10 2.813 100 the pyroauritegroup. Indeed, this line and other 001 5 2.858 2.865 021 3.s65 15 2 1q1 10 reflections appear with much greater intensity in s 2.680 2.686 310 3.393 40 2.300 5 30 2.6L9 2.619 tzl 3.312 10 2,29r 5 diffractometerpatterns, where preferred orientation 30 2.405 ?,399 401 3.250 5 2.283 30 60 2.362 2.362 303 2.L96 l0 would naturally produce such a result. 20 2.327 2.32A 4oz 2.980 10 2.r01 5 The single-crystaldata show that there-isa pro- 10 2.268 2.266 22I 2.954 60 2.@2 l0 5 2.225 | 2.234 400 2.901 5 1.940 10 nouncedsubstructure having B : 6.2312A.As re- L2.233 321 2.832 5 r.919 30 s 2.1?0 2.L20 322 2.767 20 r.900 flections with K : 2N + 1 are exceptionally weak, 2 2.074 2.073 403 2.690 s 1.762 and as the crystal was very small, we are unable to 60 1.915 I r.srz 023 2.6t0 5 t.720 L 1.910 222 2.600 5 r.061 determinewhether or not there is an extinction rule 2 L.823 1.824 004 t \71 t\ 1.590 I 1.796 1.792 503 1.4J5 l5 1.581 for 0k0 reflections. We were able to conclude that 20 L.724 t.726 2.482 5 1.570 the lattice is primitive, however, and that there is no 20 1.6,44 1.643 r24 2,408 30 1.541 I 1.s76 I r.578 607 2.287 5 1.500 glideplane. In addition,the crystal was twinned by L1.574 024 2.247 40 40 L.547 | 1.rlv 520 2.2?2 I0 reflectionon {001}. L 1.540 427 In summary we conclude that the above data 5 i.518 1.521 14r 2.195 20 20 1.458 I r.4s9 005 2,190 10 imply that "cuprohydromagnesite" and "cuproar- L 1.459 422 2.I4L 5 5 r.426 | 1.426 242 2.085 10 tinite" do not exist as they were described by | 1.425 lzl 2.050 30 Lr.42s 622 2.Q29 15 Oswald and Crook (1978).There is no direct evi- 10 1.404 1.405 403 2.018 s dence for their occurrence, and all specimensof 5 1.356 | 1.354 224 L 1.356 703 them that have been studied, including the type 20 1,302 r.302 206 20 1.290 I r.290 525 specimen,have been shown to be nakauriite. We r.387 406' L therefore urged that they be discredited. It is espe- lThis pattern is typical of all of thdse said to be of "cuprohydro- [agnesite" or "cuproartinite" specirens. cially significantthat Oswald was also in agreement zoswaldasd Crook (1979) (personalcommunication, 1980)with this recom- PEACOR ET AL.: NEW DATA ON "TEXASITE." ETC. mendation.He has informed us that in view of our Recommendationsto the I.M.A. extensivedata he can only concludethat "cuproar- It is not unusualfor a mineralto be discreditedor tinite" and "cuprohydromagnesite"are not valid to have the status of some species questioned. speciesand that all specimensof these materialsare However, our recommendationto simultaneously apparentlynakauriite or a structurally closely relat- discredit five individual, apparently well-defined ed phase.The I.M.A. Commissionon New Miner- mineralsmay be without precedent.This is in part als and Mineral Names has therefore discredited becauseof the inherent difficulties in characterizing "cuprohydromagnesite" and "cuproartinite" as the mineralsdescribed herein. It is alsobecause it is valid minerals (Kato, written communication, difficultto prove, beyond a shadowof a doubt, that 1981). a mineraldoes not exist as a valid speciesonce the Sourcesof published data cloak of authenticity is granted to it by acceptance We are unable to determinethe sourceof much of by the I.M.A. Commissionon New Minerals and Crook's data as it is presentedin the papers dis- Mineral Names.The authorsrecommend that some cussedabove. It is generallyacknowledged that the revisions be madein current proceduresused by the author of a paper will indicate which laboratories I.M.A. for accreditationof new minerals.We have and facilitiesserved as the sourceof data in one of two suchrecommendations, each of which could be threeways. First, the author's institutionaladdress implementedwithout causing considerableextra may generally be inferred to be the direct source, effort. They are: when no other reference is given. Second, it is (1) A type sampleof eachnew mineral shouldbe commonpractice to include such referencesunder required(rather than recommended)to be deposit- "acknowledgments" where laboratory facilities ed in a suitable national collection, such as the other than the author's own are utilized. Third, and SmithsonianInstitution. This sample should be less commonly, referenceto specific laboratories, clearly in matrix in order that synthetic materials equipment,etc. may be found at appropriate places might be guardedagainst. It should be large enough in the text, especiallyunder sectionswith titles such to act as a source for duplicating the determination as "Procedures." These or other references to ofcharacteristicproperties. That is, there shouldbe sourcesof data are generally not clearly defined in sufficient material to provide electron microprobe the papersin question. and X-ray diffraction data, at the very least. If so For example,in the paper describing"cuprohy- little of a proposedmineral exists that this cannotbe dromagnesite" and "cuproartinite" we are unable done, then the mineral should receive speciesac- to determine the sourcesof any of the data, includ- creditation only under the most unusual circum- ing X-ray, electronmicroprobe, optical goniometer, stances. DTA, TGA, etc. Although the addressof the au- (2) The completed paper, preparatory for publi thors is given as Mobil Oil Corporation, we have cation, should be submitted in its entirety to the determined that at least single-crystal X-ray and I.M.A. Commission, not just an abstract. This electron microprobe equipment was not available shouldbe studiedby at least selectedmembers of through that source. Oswald has informed us that the commission while taking into consideration he doesnot know the sourcesofthese data(person- both the large amount of work expended by com- al communication, 1980)and there aie no acknowl- mission members and the problem of translating a edgments.There are many similar examples in the full paper into many languages.This simultaneously paperswhich we have cited above which are much solves many problems: the article once evaluated too numerous to list here, but this single example by the I.M.A. Commissionmight actually appear will suffice to illustrate our concern. faster,eliminating the present,not.uncommon cas- It is important that the nature of the laboratories, es wherea mineralis accreditedbut is not described facilities,etc. used as a sourceof data be properly in press for several years. More importantly, the identified.Only in this way can the quality of the completepaper is crucial for a fufly informed evalu- data be correctly evaluated.This is especiallysig- ation of the mineral species.An abstract alone may nificantin the casesdiscussed in this paper,in view contain only a formula and only a very brief de- of the substantial questions about the validity of scription of the methods of analysisor sourceof the data whose sourcesare not documented.The lack laboratory. The X-ray data presentedin an abstract of such identification of sourcesof data by Crook in may have insufficient detail to allow proper infer- his publishedpapers is thereforea seriousomission. encesto be made. A completemanuscript should r69 help prevent invalid materials being given species Crook, W. W. il (1978b) Tveitite from the Barringer Hill accreditation. district, Texas. The Mineralogical Record, 9, 387. (1978) We believethat thesetwo recommendationswill Crook, W. W. il, Ewing, R. C., and Ehlmann, A. J. Rowlandite from the Barringer Hill rare-earth pegmatite dis- considerablyreduce the number of discreditations trict, Llano and Burnet Counties, Texas. American Mineral- necessaryin the future. They are simpleand do not ogist,63, 754J56. require extensive modifications in current proce- Crook, W. W. III and Jambor,J. L. (1979)Nickelbischofite, a dures. No system will ever be absolutelyperfect, new nickel chloride hydrate. CanadianMineralogist, 17, 107- but where procedures are easily implemented, and 109. Crook,W. W. III and Marcotty, L.-A. (1978)Albrittonite, a new where they may have a significant,positive effect, cobalt chloride hydrate from Oxford, Llano County, Texas. their implementationshould be seriously consid- American Mineralogist, 63, 410-412. ered. We emphasizethat these are only our own Crook, W. W. ilI (1979) Yttomicrolite, a new mineral, and a personal recommendationsand that they have not redefinition of hjelmite. American Mineralogist, U, 890-892. (1906) been considered by the I.M.A. Commission on Groth, P. Chemische Kristallographie l. Engelmann, Leipzig. New Minerals and Mineral Names. Jambor, J. L. (1975)Secondary minerals in an ultramafic intru- sion, Amos Area, Quebec.Geological Survey of CanadaPaper Note added in proof. A sample of nakauriite kindly provided by 75-14,261:263. Dr. J. Suzuki was studiedon the electronmicroprobe. It contains Kleinberg, R. (1969)Crystal structure of NiClz'6HzO at room little or no S (as is true of specimensfrom Gabbs) and thus the temperature and 4.2'K by neutron diffraction. Journal of formula of Suzuki et al. may be in error, in part. Chemical Physics, 50, 4690-4696. Liebhafsky, H. A., Pfeiffer, H. G., Winslow, E.H. and Ze- mann, P. D. (1972)X-rays, Electrons and Analytical Chemis- Acknowledgments try. Wiley and Sons, New York. We are deeply indebted to many individuals for their help and Oswald,S. G. and Crook, W. W. ilI (1979)Cuprohydromagne- cooperationin the researchfor this paper. We especially thank site and cuproartinite, two new mineralsfrom Gabbs, Nevada. Dr. J. L. Jambor,Ms. L.-A. Marcotty and Mr. S. G. Oswaldfor American Mineralogist, 64, 886-889. providing frank and invaluable data and advice. Special thanks Palache,C., Berman,H. and Frondel,C. (1951)The Systemof go alsoto Dr. M. H. Hey, Mr. J. Jones,Dr. R. L. Freed,Dr. J. Mineralogy, Vol. II, Seventh Edition, John Wiley and Sons, Haschke,Dr. E. Foord, Mr. P. Dunn, Dr. M. F. Fleischer,Dr. New York. David Garske,Dr. P. E. Brown. Dr. S. R. Bohlen. Dr. J. W. Shima, M. (1957)A new sublimate containing nickel found in a Valley, Dr. D. Alexander,Dr. D. Perkins,III and Dr. Wm. C. fumarole ofan active volcano. Joumal ofthe ScienceResearch Kelly. Many others, too numerousto name here, have provided Institute(Tokyo), 51, ll-14. encouragement,advice and data and to them we express our Suzuki,J., Ito, M. and Sugiura,T. (1976)A new coppersulfate- very sincere appreciation. carbonate hydroxide hydrate mineral, (Mg,Ni,Cu)s(SO+)+ (CO3XOH)6'48H2O,from Nakauri, Achi Prefecture, Japan. References Journalof the JapanAssociation of Mineralogy, Petrology and Economic , 71, 183-192. Crook, W. W. ilI (1976)Yttrocrasite from central Texas. The Swanson,H. E., McMurdie, H. F., Morris, M. C., Evans, MineralogicalRecord, 7, 182-183. E. H. and Paretzkin, B. (1974) Standard x-ray difraction Crook, W. W.lll (1977a)Texasite, a new mineral: the first powder patterns. National Bureau of Standards Monograph exampleof a differentiatedrare-earth species.American Min- 25. SectionII.42. eralogist,62, 1006-1008. Weast, R. C. (1967)Handbook of chemistry and physics. The Crook, W. W. III (1977b) New data on yttrocrasite from the ChemicalRubber Co.. Cleveland,Ohio. Clear Creek pegmatite, Burnet County, Texas. American Winchell, A. N. and Winchell, H. (1964)The Microscopical Mineralogist,52, 1009-l0ll. charactersof artificial inorganic solid substances.Academic Crook, W. W. III (1977c)The geology, mineralogy,and geo- Press,New York. chemistry of the rare-earth , Llano and Burnet Counties,Texas. Master's Thesis, Departmentof tieology and Mineralogy, The University of Michigan, Ann Arbor (rescind- ed in 1980). Crook, W. W. ilI (1978a)Texasite from Colorado. The Mineral- Manuscript received, October 30, 1980; ogical Record, 9, 251-252. acceptedfor publication,April 13, 1981.