222 Penfield and Sp&rry — Composition of Howlite. 



stituent, B„0 3 , being determined by difference. The air-drj^ 

 powder of the mineral lost water slowly at 100° C. but soon 

 came to a constant weight at 150° C. losing 0*83 per cent, the 

 mineral suffering no further loss by heating to 360°. The loss 

 of 0*83 per cent corresponds closely to O90 per cent, the 

 amount of water contained in the 4*32 per cent of g3'-psum 

 which the anatyzed portion contained. Thus the loss of water 

 at 150° C. as well as the S0 3 determination and the occurrence of 

 the nodule in massive gypsum suggest that there was a slight 

 impurity of gypsum in the material which was anatyzed. The 

 fact that the mineral does not part with its water at 360° C. indi- 

 cates that the hydrogen is very firmly united in the molecule, 

 probably as h} 7 droxyl. The mineral is therefore a very acid 

 silico-borate and its formula may be written H 6 Ca 2 B 6 Si0 14 , 

 differing essentially from the more common silico-borates of 

 calcium, datolite and danburite, in containing much less silica 

 and more boracic acid, causing the mineral to be classified 

 more naturally with the borates than with the silicates. 



On igniting the mineral in a closed tube water is given off 

 which reacts strongly for boracic acid with, tumeric paper ; the 

 water in the above analysis was therefore obtained by ignition 

 and absorption in a chloride of calcium tube. The boracic acid 

 was determined by the admirable method suggested by Prof. 

 F. A. Gooch.* The other constituents were determined by 

 the usual methods. 



2. Note on the Oooch method for the determination of boracic acid. 



Before commencing the above analysis we carefully reviewed 

 the method described by Prof. Grooch,f making a number of 

 determinations of B a 3 in borax in order to become fully 

 acquainted with the method. The distilling apparatus de- 

 scribed by him proved ver\r satisfactory. The boracic acid in 

 our experiments was set free by nitric acid, distilled over with 

 methyl alcohol and evaporated with a weighed quantity of 

 ignited calcium oxide. Without the use of very large plati- 

 num crucibles the evaporation of the boric ether with slaked 

 lime is attended with some difficulty. The calcium borate 

 together with the excess of lime form as a thick crystalline 

 precipitate at the bottom of the crucible and if the methyl 

 alcohol boils during the evaporation some of the precipitate 

 will be carried mechanically out of the crucible; then too the 

 very- strong ignition of the CaO necessary to bring it to a con- 

 stant weight is not so readily accomplished if the crucibles are 

 very large. The following method was found to give very 

 good satisfaction and can be used by any one having an ordin- 

 ary sized platinum crucible and preferably a platinum dish. 

 * Am. Chem. Journal, ix, p. 23. f Loc. cit 



