266 THE EEV. J. MAGENS MELLO, M.A.^ E.G.S., ETC.^ 



unnecessary. If, as Mr. Mello suggests, " accidental fracture " 

 were the only factor concerned, it might be so ; but what has 

 been urged by M. Arcelin (p. 239) from his own observations, as 

 to the operation of other natural agencies, strengthened by the 

 observations of Dr. Livingstone, the Marquis de Nadaillac, M. 

 Lepsius, and others, shows that that is not the case. Besides the 

 action of forest fires in splitting flints, we have to allow for the 

 sun's heat ; and it so happens that I have in my possession 

 specimens of a variety of tabular flint, obtained by my former 

 pupil. Captain H. Gr. Lyons, R.E., F.G.S. (now on the staff of the 

 Egyptian Army at Wady Haifa), during a i*ecent tour in the 

 desert between the Nile and the Red Sea, which have undergone 

 fracture on the upper and exposed side only (as the specimens lay on 

 the desert sands), and that to such an extent that the whole sur- 

 face is in several specimens converted into a series of saucer-like 

 depressions, some of which contain a little mammilla-like elevation 

 in the centre. Similar specimens were, I believe, given by Capt. 

 Lyons to Professor Judd. If, as seems most probable, such forms 

 were produced by the alternations of powerful heating during the 

 day by the sun's radiation, and the rapid cooling which results at 

 night-fall from the powerful terrestrial radiation, which takes 

 place always under the clear sky of the desert, it is more than likely 

 that cracks Avould be formed in this manner, and that little rough 

 segments of spheres easily broken would in this way flake off. A 

 true flint, it must be remembered, is more or less coTnposed of 

 colloid silica, and such silica (as my own observations recorded 

 elsewhere show) does actually assume, in some cases, the cha- 

 racter of a glass; the molecular structure of which lends itself 

 readily to the formation of shrinkage-cracks. This is as true of a 

 silica-glass formed by an aqueous process, as it is of a vitreous 

 rock (like perlite) produced by what are commonly known as 

 igneous agencies. 



Again, it is a matter of observation, that when a crack has been 

 once started, whether by a blow or by molecular action, iron under 

 favourable conditions may be carried in solution into it by natural 

 solvents, and that with subsequent desiccation, this iron may 

 undergo oxidation into the peroxide by taking up atmospheric 

 oxygen, with a corresponding inci'case of bulk. The molecular 

 forces set up in this way must act as a powerful wedge to widen 

 and extend even the closest fractures. I have in my possession a 

 flint pebble from the Reading beds which shows this. Flints, too, 



