May 4, 1906.] 



SCIENCE. 



705 



circular eminences, averaging, say, twenty feet 

 in diameter, seldom exceeding two feet in 

 height, occurring thickly studded over exten- 

 sive areas of both forest and prairie lands. 



For nearly twenty years I have been ob- 

 serving these mounds in the second bottoms 

 of the rivers of the southern coastal plain, 

 the coast prairies of Louisiana and Texas, in 

 southern Arkansas, Indian Territory and in 

 the extension of the old valleys of the rivers 

 toward the great plains. 



I do not believe that they owe their origin 

 to the uprooting of trees, the handiwork of 

 man, to glacial agencies, or the pressure of 

 underground gases, as some of your corre- 

 spondents and others have alleged. 



As to Mr. Farnsworth's theory that they 

 are the result of uprooting trees : while I have 

 seen many mounds in the forests which 

 have thus been made, this hypothesis can not 

 apply as a general explanation, owing to the 

 fact that millions of the mounds occur on the 

 newly made coast prairie of the Texas region 

 which is not and has never been inhabited by 

 forest growth. Two weeks ago I drove 

 through thousands of these mounds on the 

 mainland of Texas opposite Galveston Island, 

 and any one familiar with the conditions of 

 that portion of the coast prairie will imme- 

 diately abandon the uprooting tree theory. 



Mr. Bushnell's theory that these mounds 

 were made by man is also totally inadequate. 

 I am well aware of the fact that in poorly 

 drained ilat areas of the old flood plains and 

 second bottoms of the Mississippi Valley and 

 its tributary laterals, mounds do exist which 

 were constructed by aboriginal hands, but the 

 mounds under discussion are not of this class. 

 I have seen hundreds of these dissected by 

 roadways and other cuttings and they show 

 no trace of human work. Furthermore, they 

 are so numerous and extensive that their con- 

 struction by men would have required a larger 

 population than has ever yet inhabited these 

 regions, or than it could possibly support. 



The glacial theory can also be dismissed 

 with the statement that in most instances 

 these mounds inhabit non-glacial formations. 



Another theory, which has probably been 

 mentioned by some of your correspondents, 

 is that the mounds are produced by the 

 ascending gas above oil pools, has wide 

 local adherence and should be discountenanced 

 because of its economic misapplication. In 

 fact, the name ' gas mound ' now being used 

 locally by the people for this class of phe- 

 nomena in Texas is the only specific one which 

 I have heard applied. Many charlatans and 

 even misguided honorable men are holding to 

 the ' gas-mound ' theory and express astonish- 

 ment when any one disputes it, notwithstand- 

 ing the fact that the identical mounds are 

 found to occur in many districts where not 

 the least sign of oil or gas has been discovered. 

 This theory has cost useless expenditure of 

 many thousands of dollars in drilling for oil 

 pools. 



While frequently feeling, like Professor 

 Forshey, as quoted in Mr. Farnsworth's article, 

 that the more familiarity I have with these 

 mounds, the less explicable they seem to me, 

 I am of the decided opinion that they are 

 natural products of certain topographic, cli- 

 matic and geologic conditions. 



It has been my observation that the mounds 

 always occur upon areas of poorly drained 

 sublevel surfaces in regions of abundant, 

 periodic rainfall. They are also always un- 

 derlain by formations of alluvial materials of 

 relatively uncompact sands and clays. 



The rainfall upon these places, owing to 

 absence of well-defined rimways, stands until 

 it is evaporated or absorbed. The materials 

 have different capacities for absorption and 

 transmission, retention and loss of water, re- 

 sulting in the unequal settling of the ground 

 and the formation of the mounds and their 

 interspaces. In the dry season these mounds 

 are frequently augmented in size by drifting 

 sands. 



While writing upon this subject, I might add 

 that last year I observed mounds exactly sim- 

 ilar to those of the southern coastal plain region 

 upon the Bavicora plain near the top of the 

 Western Sierra Madre of Mexico at an alti- 

 tude of nearly 7,000 feet. This plain, like the 

 coast prairie, is an extensive flat upon which 



