iSoS-] 477 [Criswold. 
may infer that tlicy formerly had a considerable extent over the 
now exposed Palaeozoic rocks, and the marked discordance of 
streams to structnre north of the Cretaceous border seems to 
demand an origin fi-om superimpositioni, thus supjwrting the 
inference. North of the present Cretaceous border we now find 
the stripped Palaeozoic rocks presenting a very even surface, but 
slightly dissected by streams, and rising gradually to the north. 
The farther north this surface is traced, therefore the longer 
uncovered, the more it is found to be cut up by streams and the 
more dependence of the streams upon the geological structure is 
noticed. At a distance of fifteen to twenty miles from the border 
the plain as such is no longer seen, but we have the projected 
position of the plain surface indicated by the uniformity in the 
elevation of the crest lines of the ridges. 
The mountain system in Arkansas above noted is believed l>y 
Branner- to have been an outlier of the Appalachian System 
which he considers to have extended farther south across the 
Mississippi endmynient into Texas and perha})S farther. Then 
the mountain system thus constructed was worn down to the 
peneplain stage and partly buried under a Cretaceous cover. A 
similar relation of Cretaceous to a denuded Palaeozoic land in 
central Texas as shown by Tarr^ evidences a further westward 
extension of similar conditions. 
As the Silurian strata of central Arkansas by change in the 
character of the sediments indicate a derivation of materials from 
the south, the supjiosition of an old land mass in that direction, 
extending from the known old lands of Alabama to those of 
central Texas and bordering the extended and now buried 
Appalachians on the south, is not unreasonable. This would 
give not only a complete barrier to southward-llowing drainage, 
but would send the drainage of Palaeozoic and ])Ost-Palaeozoic 
time northward just as in the Virginia-Carolina region it flowed 
northwestward. It may be urged that such a condition is purely 
hypothetical and therefore of no value ; still, as may be noted on 
1 Opus cit., p. 218-222. 
2 Ann. rept. geol. surv. Ark., v. 3, ]). 213, 1890. 
■'Notes on the physical geography of Texas. R. S. Tarr. Froc. acad. nat. sci. 
I'hil., 1893, p. 317-318. 
