Chap. IV. ] 
PSILOMELANE. 
99 
especially at Yeruli. The best example of this mineral I have seen 
showing a reniform shape is a specimen foimd loose at Garbham in 
the Vizagapatam district (see fig. 2, Plate 5). At Gugiildoho I found a 
very fine specimen of psilomelane with a form that can be best described 
as liver-shaped or hepatiform. The specimen as first found was some 10 
inches across and resembled nothing so much in shape as the liver of one 
of the large carnivora, such as a panther. It was made up of concentric 
layers and is the specimen No. 1157 of which an analysis is given on page 
100. At this locality is to be foimd almost every variety of 
concretionary structure exhibited by psilomelane, there being many 
fine examples of mammillary forms. Fig. 2, Plate 6, shows one of 
these. The best example of stalactitic psilomelane I have seen from 
India was found by Mr. Geeson at Garbham in the Vizagapatam district 
(see fig. 1, Plate 5). According to Dana the lustre varies from sub- 
metallic to dull. This applies to most of the Indian psilomelanes, but 
there is one variety, found at Garbham and Avagudem in the Vizaga- 
patam district, and at Kumsi and other deposits in Mysore, that on fresh 
fracture resembles very closely in colour and lustre a piece of metallic 
lead, so that its lustre can be described as mecallic. This is the variety 
of which an analysis is given on page 100 (No. A. 372). The streak 
of this mineral is said by Dana to be brownish-black in colour. Althoug h 
this undoubtedly holds for some examples, yet the Indian psilomelanes 
have, as often as not, a black streak. In colour the Indian specimens 
vary from iron-black through dark steel-grey to an almost bluish-grey, 
this bluish tint being especially noticeable when one is collecting 
specimens in the blazing sunlight of an Indian dry- weather day. 
It win be noticed that the specific gravity of the specimens of which 
analyses are given on page 100 all lie between the 
Specific gravity, j^jj^^^-g gjyg^ }jy Dana. They tend, however, to be 
almost invariably over 4 in value, the only exception being the lead- 
like variety. 
I have already mentioned in discussing the composition of hoUandite, 
which is to be regarded as the crystalline form of 
Composition. mineral, that analyses of both minerals con- 
form very closely to the formula first suggested by Laspeyres, so that 
both psilomelane and hollandite can be regarded as derived from a 
hypothetical acid of the composition H4Mn05. Below I give a series 
of five analyses carried out by Messrs. J. and H. S. Pattinson of New- 
castle-on-Tyne on specimens collected by myself. For the purposes 
of analysis I picked out pieces that were as homogeneous as could be 
I H 2 
