59 
If the percentage amount of manganese dioxide be calculated 
on the dried specimen, it is found to amount to 82 '21. 
Traces of zinc and of sodium were also found. 
The specimen of native silver, the analysis of which is 
appended, was sent to me from the Thames Goldfield, New 
Zealand. It presented the appearance of arborescent crys- 
tals radiating outwards from a central mass at acute angles. 
A very small quantity of crystalline quartz could be seen in 
the cavities of the mass. 
The silver was estimated by precipitation as chloride; 
the copper, mercury, &c., were precipitated by means of 
sulphuretted hydrogen ; the copper was separated by solu- 
tion in nitric acid, and estimated in this solution by the 
colorimetric method depending upon the use of potassium 
ferrocyanide described by Carnelley (Proc. Manchester Lit. 
and Phil. Soc., 1875-76, p. 24). The iron was precipitated 
by means of ammonium sulphide, the precipitate was dis- 
solved in dilute hydrochloric acid, and the quantity of iron 
estimated by Carnelley ’s colorimetric method (Proc. Lit. 
and Phil. Soc., vol. v., p. 346). 
The sulphide of mercury was dissolved in aqua regia, and 
an attempt was made to estimate the quantity of the metal 
by Hannay’s volumetric method (Chem. Soc. J. [2] xi. 565). 
The quantity present was, however, too small to admit of 
estimation by this method. 
The mercury was approximately estimated by comparing 
the depth of colour produced in the liquid on the addition 
of sulphuretted hydrogen water with that produced in an 
equal bulk of pure water to which a known volume of 
standardised mercuric chloride solution had been added. 
This method did not give very satisfactory results : it was 
found that the shades of colour in the two liquids were not 
exactly the same. For this reason the quantity of mercury 
stated below must be regarded as approximative only. 
