HARRISON COUNTY. 127 



higher ridges of the county (and they appear to be), it is true that the far- 

 ther we get away from the breaks along the sides of these ridges or the lower 

 lying hills along the plateau margin the fewer deposits of ferruginous gravel 

 are to be found. Another circumstance connected with the existence of these 

 quartz and other crystalline pebbles is their existence among the brown and 

 yellowish brown sands frequently found capping the highest points of the 

 central plateau and consequently above the sources of the ferruginous gravels. 



The origin and movement of these ferruginous gravels may be illustrated 

 by Section III, already given, and shown in Figure 6. In this section the 

 brown sand overlying the summit of the hill to a depth of two feet con 

 tains so much ferruginous gravel derived from the weathering of the sand 

 stone bed that the deposit might almost be considered a ferruginous gravel 

 The heavy deposit of brown sand lying at the foot of the slope shows com 

 paratively few pieces of ferruginous matter. The sand as well as the grave 

 at both places are of the same nature and both are derived from the same 

 source. 



Diagram showing relative positions of ferruginous gravel and brown sand 

 on hill two miles southeast of Marshall (Section III): 



h 



Fig. 6. 

 a, Brown sand, b, Ferruginous parting, c, Stratified blue clay, d, Dark blue stratified 



sandy clay, e, Creek. 



3. The third characteristic of this brown sand, and the most distinguish- 

 ing one, is the presence of fossil wood in large quantities. In size these frag- 

 ments vary from two to ten inches in diameter and from six inches to ten 

 feet in length. Some of the fragments indicate trees having a growth of 

 from twenty-eight inches to three feet in diameter. 



The wood is generally silicified and has a white or brown color. The 

 colors, however, vary according to the prevalence of iron in the sand. Some 

 of the fragments are stained a dark iron brown, and others have the mole- 

 cules of vegetable matter largely replaced by a hydrate of iron. Some of 

 the fragments are soft and chalky white, readily soiling the fingers, and 

 others are hard and sonorous when struck with the hammer. 



There does not appear to be any abrupt or sudden change between the re- 

 gion occupied by the brown sand and the area overlaid by the light gray 

 silty sand, forming the third division and lying to the south along the Sabine 

 River and eastward along the Louisiana State line. These gray sands occupy 



