484 Proceedings of tlhe Eoyal Society 
by winds, or may have come from the ground wdiere the ashes were 
collected. 
In spite of all the uncertainties which the exact diagnoses of 
volcanic dust present, we can consider them often, from the point 
of view of their mineralogical composition, as analogous with the 
augite-andesites. We know, besides, that it is to these rocks that 
the lavas of the volcano of Krakatoa should be referred. 
The ashes which fell at Batavia on the 27th August 1883, and 
samples of which were sent to Holland by M. Wolf, resident on 
that island, have been analysed with the following results : — 
I. 1 T19 grm. of substance dried at 110° C. , and fused with carbonate of soda 
and potash, gave 07799 grm. of silica, 07754 grm. of alumina, 0’0911 grm. of 
peroxide of iron, 0*0401 grm. of lime, 0*398 grm. of pyrophosphate of mag- 
nesia, answering to 0*01434 grm. of magnesia.* 
II. 1*222 grm. of substance dried at 110° C. gave 0*0335 grm. of loss on igni- 
tion (water, organic substances, chloride of sodium); the same substance 
treated with hydrofluoric and sulphuric acids gave 0*1161 grm. of chloride of 
sodium and potassium, and 0*0118 grm. of chloroplatinate of potassium, 
answering to 0*0118 grm. of potash and to 0*0188 grm. of chloride of potas- 
sium; by diff‘erence = 0*0973 grm. of chloride of sodium, answering to 0*05163 
of soda. 
III. 1 *7287 grm. of substance dried at 110° C. was treated in a closed tube with 
hydrofluoric and sulphuric acid. The oxidation required 2 *3 c. c. of perman- 
ganate of potash (1 c.c. =0*0212 grm. FeO), answering to 0*047876 grm. of 
peroxide of iron. 
I. 
II. 
III. 
SiOa 
65*04 
65*04 
AI 2 O 3 
14*63 
14*63 
FegOg 
4*47 
4*47 
FeO 
. . . 
2*82 
2*82 
MnO 
traces 
traces 
MgO 
1*20 
1*20 
CaO 
3*34 
3*34 
K^O 
0*97 
0*97 
Ha 20 
4*23 
4*23 
Loss 
2*74 
2*74 
99*44 
It will be understood that it is barely possible to submit this 
analysis to discussion. The abundance of vitreous particles in the 
ashes renders illusory the calculation of the values obtained, and 
* A recent determination of titanic acid has given 0*62 per cent. TiO. 2 . 
