376 & L. Penfield — Spangolite, a new copper mineral. 



of 26° 25 7 for the extraordinary and 28° 46' for the ordinaiw, 

 measuring from the brightest part of the spectra, which from 

 the spectroscopic examination we have located at about X 525, 

 and we obtain for the two indices of refraction co 1*694, e 1*641. 



Other physical properties. — The hardness of the mineral on 

 the basal plane is about 2, on the pyramidal faces nearly 3. 

 The specific gravity was taken very carefully with a chemical 

 balance on the purest material, which was selected for chemical 

 examination. After boiling the crystals in water, to expel any 

 air, three separate portions weighing respectively 0*2143, - 1787 

 and 0*1538 grams gave 3*147, 3*133 and 3*142, an average of 

 3*141 as the specific gravity. 



Chemical composition. — More than three grams of excep- 

 tionally pure material were readily obtained by sacrificing about 

 one-half of the crystals on the specimen. As the material was 

 somewhat limited a method of analysis was adopted by which 

 nothing could well escape detection and a qualitative and quan- 

 titative analysis w T as carried on with a single sample, the results 

 of which are given below. The fourth analysis was made on 

 an entirely different sample from that which yielded the figures 

 in the first three columns. 





I. 



II. 



III. 



IV. 



Average. 



Calculated 



for Cu 6 Al 



Ratio. ClSO ]0 9H 2 O 



SOf 





10-08 



10-11 



10-14 



10 11 



•126 



1-01 



10-03 



CI 



4-12 







410 



4-11 



4-11 



•116 



0-93 



4-45 



A1 2 3 .... 



6-59 







6-51 



670 



6-60 



•064 



0-51 



645 



CuO ." 



5957 



- - - - 



59-47 



59-50 



59-51 



•7495 



6-00 



59-75 



H 2 







20-32 







20-49 



20-41 



1-134 



9-07 



20-32 





10074 



101-00 







O, equivalent 



of CI... 



•92 







1-00 



99-82 ' 



100-00 



The analysis yields a ratio of S0 3 : CI : A1 2 3 : CuO : H 2 = 

 1*01: 0*93: 0-51: 6*0: 9*07 or very nearly 1:1:*5:6:9, from 

 which we obtain a rather complicated and remarkable formula, 

 Cu r A1C1SO 10 , 9H 2 0. No doubt can however exist concerning 

 this formula; not only was the material beautifully crystallized 

 and of unusual purity, but the analysis of two separate samples 

 are identical within the error of analysis and the ratio is through- 

 out very sharp ; moreover the calculated composition agrees 

 very well with the results of analysis. A slight deficiency in 

 chlorine may result from a partial replacement of that element 

 by hydroxyl, which, if true, would diminish somewhat the 

 slight excess of water. The method of analysis was as follows : 

 A weighed quantity of the mineral lost water slowly by stand- 

 ing in a desiccator over sulphuric acid, amounting to - 30 per 

 cent in thirty six hours, but it regained almost all of this loss 



