224 



To untangle this mass is an enormous labor, and one perplexing in the 

 extreme. Many have undertaken it, and have given it up in despair. In 

 1874, Mr. J. T. Gardner, then connected with this Survey, undertook a 

 part of this labor, and succeeded in carrying it through with a consid- 

 erable degree of success. His problem was, to obtain the true eleva- 

 tion of Denver, Colo., which serves as a hypsometric base for our work 

 in Colorado. To this end, he connected a number of lines from the 

 Atlantic and Gulf coasts, rooted out nearly all the errors, and finally 

 obtained results for the heights of several of the most important rail- 

 road-centers, which, from their close agreement, show a strong proba- 

 bility of being correct. Should these results, on the application of 

 further tests, prove trustworthy, they will form a frame-work, by use of 

 which the problem will be simplified very much. 



Elevations measured by barometer are liable to large errors, unless 

 the work is done under the very best circumstances. The most potent 

 causes of error are too great distance, horizontal and vertical, from 

 the barometric base. In the earlier days of exploration in the West, 

 it was impossible to obtain even tolerable conditions for the execution 

 of hypsometric work. Hence, naturally, the earlier measurements of 

 elevations in the West are found to iiresent a wide range of error, some 

 being changed scarcely any by later and more reliable measurements, 

 others being several hundreds of feet — indeed, in some cases reaching 

 a thousand feet — from the truth. 



In constructing these curves, though I have made use of the many 

 thousands of elevations in my j^ossession, I have borne in mind that ele- 

 vations of isolated points merely, without a knowledge of the surround- 

 ing topography, are of little use. Indeed, in many cases they are worse 

 than useless, a&they may tend to mislead. 



The mean heights of mountain-ranges have been expressed as nearly 

 as possible. They will, no doubt, be subject to much criticism, as many 

 of them are based on insufficient information, and my estimate of the 

 mean heights of others, based on the heights of a few peaks, may not be 

 correct. 



In the canons of the Colorado and Green Rivers, I have thought best, 

 owing to the small scale of the map, to run the contours into a single 

 line. 



The sources of information consulted and made use of in this work 

 are almost numberless. I will enumerate the principal ones. For a 

 knowledge of the Appalachian system and the country adjacent, I am 

 indebted mainly to the article by Professor Guyot on the Appalachian 

 system, published in vol. xxxi (1861) of Silliman's Journal; also, in 

 various State geological and geographical reports, there is much ma- 

 terial of value concerning the Eastern States. Information concerning 

 the valley of the Mississippi is scattered far and wide, and to specify 

 authorities would fill a volume. Reports of State and General Gov- 

 ernment surveys and profiles of railroads have supplied the greater part , 

 of this information. 



On the slope of the great plains, between longitudes 96° and 104°, 

 the contours are controlled largely by the plotted heights. These 

 heights have been collated from, first, the profiles of the several rail- 

 roads which cross the plains, the Northern- Pacific, Union Pacific, Bur- 

 lington and Missouri River, Kansas Pacific, Atlantic and Pacific, Ar- 

 kansas Valley, and others of less importance; and, second, from the 

 profiles of the routes of the numerous expeditions which have crossed 

 the plains in various directions. 



The Llano Estacado and the western part of Texas are known mainly 



