824 



SCIENCE. 



[N. S. Vol. XXI. No. 543. 



larger numbers and worked more industri- 

 ously, for the present ant-hills, are diminutive 

 as compared with the mounds. 



For some time the writer has entertained a 

 theory very similar to that mentioned by Pro- 

 fessor Branner. As above stated, these 

 mounds are always on clay soil. In the Paleo- 

 zoic region of Arkansas, they are on residual 

 clay soils only a few feet deep, and of shale 

 origin. As stated, the drainage where they 

 occur is usually poor. These facts point to 

 the action of ground-water within the clays 

 or shales as being in some way responsible for 

 the mounds. The action is thought to be one 

 of the segregation of mineral matter, or as 

 Professor Branner puts it, ' concretionary ac- 

 tion on a large scale.' After the segregation, 

 the volume may be further increased by hy- 

 dration, oxidation and other chemical changes. 



This idea was fii'st suggested by a section of 

 one of these mounds in the Arkansas valley 

 that was brought to view by a small stream 

 having cut its way through it and into the 

 shale below, as shown in the figure. The un- 

 eroded portion of the mound was typical of 



Section of a natural mound cut through by a 

 stream. 



the hundreds in the vicinity, and the general 

 conformity of the surface to the arch of the 

 shale would lead one to believe that the mound 

 was due to the lifting of the shales beneath. 

 While the writer has seen many sections 

 through these mounds, this is the only one 

 that discloses the shales, so that its value lies 

 only in its suggestiveness. 



In the Paleozoic region of Arkansas these 

 mounds occur on at least three different beds 

 of shale, two of which belong to the Lower 

 Carboniferous, and the remaining one or more 

 to the Coal Measures. These are all carbon- 

 aceous, clay shales. If their cause should 

 prove to be a chemical one, induced by the 

 action of ground-water, the question will pre- 

 sent itself as to why they do not have a wide 



geographic distribution as well as a geologic 

 one. The explanation would probably be 

 found in the climatic coiiditions where the 

 mounds occur. But this is scarcely worth 

 speculating on till the origin of the mounds 

 is determined. A. H. Purdue. 



The U^-IVERSITY of Arkansas, 

 Fayetteville, Ark., 

 April 10, 1905. 



THE BASALT MOUNDS OF THE COLUMBL\ LAVA. 



The recent discussion of various types of 

 mounds of uncertain origin leads me to call 

 attention to a form common in eastern Wash- 

 ington, which seems thus far to have escaped 

 printed notice. Very conspicuous examples 

 are found in the vicinities of Spangle and 

 Medical Lake. Similar ones occur near Win- 

 ona in the old bed of the Palouse River. Less 

 striking examples are generally found along 

 the crests of all the canyons hewn out by 

 streams in the basalt, especially on the north 

 walls. The general proportion of these 

 mounds is about that of an upturned saucer, 

 but occasionally more convex. The most con- 

 spicuous are about four feet high, about twice 

 the height of the more usual ones. In diam- 

 eter they vary from ten to twenty feet, or 

 rarely more. The first generalization that 

 forces itself upon one is that these mounds 

 occur only where there has at one time been 

 flowing water. They are conspicuous enough 

 even at the top of Snake River canyon, though 

 the river now flows on a bed two thousand feet 

 below. Where these mounds occur along the 

 crests of canyons there is usually but a single 

 series of them. Where, on the other hand, 

 they occupy the old beds of broad shallow 

 streams, as at Medical Lake and near Spangle, 

 there may be acres of them, rather evenly scat- 

 tered, and often quite close together. 



The soil of these mounds shows no appre- 

 ciable difference from the surrounding soil of 

 basaltic origin, and except in the rare cases 

 where water stands about their bases, they do 

 not support a vegetation more or less lux- 

 uriant than that of the surrounding soil. 

 There is nothing, in short, in the structure 

 of the ordinary mounds to give a clue to 

 their origin. 



I 



