108 Dr. F. H. Hatch— Gem- Gravel, W. Coad of Africa. 



easily removed by a preliminary treatment with bromoform of density 

 2'9. The residue still contained limonite, some rock fragments, and 

 particles of tourmaline, the bulk of which was removed by treatment 

 with a borotungstate of cadmium solution, having a density of 3 '4. 

 A mixture of heavy minerals was then obtained which could only be 

 separated bv hand-picking with the aid of a lens. This, although 

 a tedious task, was found to be quite practicable ; and a number of 

 parcels of distinct minerals were thus obtained, the final determination 

 of which was arrived at by taking the density, testing the hardness 

 and streak, and, where practicable, measuring the crystal angles. 

 In those cases where there was sufficient material, chemical tests 

 were also made, both by dry and wet methods. 



For permission to carry out this work in the Chemical Laboratory 

 of the Mineralogical Department of the University of Cambridge, I am 

 indebted to Professor W. J. Lewis, F.R.S., and for assistance in 

 making a few goniometric measui-ements of crystals, I liave to thank 

 Mr. Arthur Hutchinson, Demonstrator of ^lineralogy in the University 

 of Cambridge. 



Of the samples examined the two principal ones were : A (No. 30) 

 gravel from the upper flats, and B (jS"©. 31) from the lower flats. 

 Sample A consisted of kyanite, corundum, pyrites, ilmenite, zircon, 

 rutile, epidote, and gold. Sample B consisted of garnet, kyanite, 

 pyrites, chroraite, magnetite, haematite, zircon, and diopside. It will 

 be observed that in sample B from the lower flats the minerals garnet, 

 chromite, magnetite, haematite, and diopside occur, which are not 

 represented in the sample from the upper flats ; and, moreover, the 

 garnet occurs in great abundance. It is obvious, therefore, that some 

 source of supply was available for the formation of the gravels of the 

 lower flats which did not exist in the upper reaches of the river. 

 In all probability this was the garnetiferous gneiss already mentioned. 

 The presence of gold in sample A was accidental, as the gold from 

 both samples had been previously removed. The particle of gold 

 found in the sample was coated with amalgam, and its true character 

 only became apparent after driving off the mercury by heat. 



Jbiamond. — Of the uncut diamonds still in the London office of the 

 Liberian Company and examined by me, some are perfect octahedral 

 crystals showing the usual rounded edges and striated faces due 

 to vicinal forms. Others are irregular cleavage fragments showing 

 no original crystal faces, while others again are partially bounded by 

 original faces and partly by octahedral cleavages. There are also 

 flattened rudely triangular forms (macles) due to twinning on the spinel 

 type. In many cases the faces are smooth with occasional triangular 

 pittings ; and, as a rule, there is no sign of true abrasion. Some 

 of the faces of the crystals, however, have a roughened, fretted, or 

 frosted surface which must, I think, be attributed to corrosion while 

 the crystals were still in the igneous magma in which they were 

 formed, rather than to any subsequent attrition. Many of the stones 

 are pure white and flawless ; but some have a yellow or brown 

 tinge and there is the usual proportion of spotted material. The 

 average value of the whole parcel has been appraised by A. E. jS^orth 

 and Co., of Hatton Garden, at £2 lis. per carat. 



