938 
MANGANESE DEPOSITS OF INDIA : 
DESCRIPTIVE. [PaKT IV: 
in the amount of manganese. This is of course what one would have ex- 
pected, except that it is not so great as the character of the rejected 
material suggested. Sample 32 should however have shown a decided 
improvement in analysis over that of sample 31. But the figures show 
very little difference between the two. In preparing the sample for 
analysis duplicate lots of the final powder were bottled. Thinking that 
a mistake had been made in labelling, and that the two samples sent 
for analysis as 31 and 32 were either both 31 or both 32, I sent one of 
the remaining samples to Messrs. J. and H. S. Pattinson of Newcastle 
for analysis. The result is shown below : — 
Sumph No. 32. 
Manganese peroxkk- ....... 57'27 
Manganese protoxide . . . . . . . 19'66 
Ferric oxide ......... 9-86 
Alumina .......... 1'58 
Baryta 0-53 
Lime .......... 1'15 
Magnesia 0 43 
Potash 0-73 
Soda 0-17 
Combined silica . . . . . . . . 4*20 
Free silica 0-25 
Sulphur 0-025 
Phosphoric oxide . . . . . . . . 0-179 
Arsenic oxide ......... 0-066 
Cobaltous oxide ........ Nil. 
Nickelous oxide ........ Nil 
t!upric oxide ......... Trace. 
Lead oxide ......... Nil. 
Zinc oxide ......... Nil. 
Titanic oxide ......... 0-08 
Chlorine Trace. 
Fluorine Nil. 
Combined water ........ 3-00 
Moisture at 100° C 0-70 
Carbon dioxide ........ Nil. 
99-880 
Manganese 51-44 
Iron 6-90 
Silica (total) . . , 4-45 
Phosphorus . ...... 0-078 
This analysis is remarkably close to that of sample 32 carried out at 
the Imperial Institute, and shows that no mistake was made in the bottling 
of the samples. 
Taking the analyses of samples 32 and 33, it is evident that if they 
were mixed in the proportions of the v;eights of ore to which they corre- 
spond, namely 3 to 2, the analysis of the original sample would not be as 
good as that shown for sample 31. Taking only the four constituents 
