﻿composition of Herderite and Beryl. 109 



from the same material. After almost completing our analysis 

 and rinding only 5*27 per cent of fluorine we tried the follow- 

 ing experiments. A little herderite powder was placed in a 

 hard glass tube, sealed at one end and ignited ; at first only a 

 slight film of water condensed in the cold part of the tube ; 

 by strong ignition over the blast lamp, however, there was a 

 sudden evolution of hydrofluoric acid, which etched the glass 

 very perceptibly near the mineral and deposited a film of silica 

 and very acid water, as marked as in an ordinary reaction for 

 fluorine in a closed tube with acid sulphate of potash. In our 

 experience we have never seen any hydrous fluoride which 

 gives off such strongly acid water and such a marked fluorine 

 reaction. Some of the powdered mineral was placed in a 

 closed glass tube, covered with a layer of dry sodium carbonate 

 and strongly ignited; neutral water was given off and con- 

 densed as a ring, which indicated more than a trace of water. 

 The water was obtained in the following way, About two 

 grams of calcite were ignited over the blast lamp till a con- 

 stant weight was obtained. The mineral was then weighed 

 into the same crucible, the lime was slaked with water, the 

 contents of the crucible were carefully dried and then ignited 

 till constant weight was obtained, the water being calculated 

 from the loss of weight. The slaking of the lime makes an 

 intimate mixture of the mineral with the lime, and a prelimin- 

 ary experiment proved to us that only neutral water was 

 driven off. In I there was a slight mechanical loss in slaking 

 the lime, which caused the water determination to be too high, 

 the P 2 6 too low. In III the water was obtained from a larger 

 quantity of mineral. No sublimate was formed on the cover 

 of the crucible by the volatilization of any fluoride. The 

 analyses were made on air-dry powder which lost 010 per 

 cent by drying for one hour at 100° C. The beryllium pre- 

 cipitates were always of a light cream color after ignition, indi- 

 cating that not more than a minute trace of iron was present. 

 The following quantities of mineral were used in making the 

 analyses: I. 4552 grams; II. 1*0029; III. P 2 5 and bases 

 •7336, F -9692, H 2 1-1612. 



I. 



II. 



in. 



Ratio. 







Calculated, 



P.O. 43-47 



.. 



43-74 



•308 





1- 



43-83 



BeO 



15-28 



15-51 



•620 





2-01 



15-44 



CaO 



33-61 



33-G7 



•601 





1-95 



34-57 



F 







5*27—38 



•138 ) 

 •205 j 



•343 



1-11 



5-86 



H 2 4-37 ? 







3-70 



2-77 





101-89 



102-47 



O equivalent of F 





2.22 









2-47 



99-67 100-00 



