July 29, 1887.] 



SCIENCE. 



51 



use of the rectangular surveys of the General Land Office, which 

 extend over this region. The township-plats supply more or less 

 fully the drainage, and, in addition to this, they cover the ground 

 with located points, the township, section, and quarter-section 

 corners. Furthermore, throughout the settled portions of Kansas, 

 and in the greater portion of Missouri, the roads, fences, hedges, 

 etc., mark the lines of subdivision in such a manner that the 

 country is graphically subdivided, and the location of features 

 horizontally becomes simply a matter of sketching. In Kansas 

 these lines of subdivision are controlled by belts of triangulation, 

 which, starting from lines of the transcontinental belt of the Coast 

 and Geodetic Survey, run westv^'ard, midway between parallels of 

 latitude. In Missouri the work is controlled by the transcontinental 

 belt of the Coast and Geodetic Survey. Each triangulation point 

 is connected with the nearest township or section corner. The 

 topographic work consists in the verification of the drainage of 

 the Land Office plats, in supplementing it wherever necessary, 

 and in adding the culture and relief. Heights are measured by 

 barometer, and the profiles of railroads are utilized. 



The work in the various fields of the Western section is carried 

 on by methods quite similar to one another. The triangulation in 

 Texas rests upon a base-line, measured near Austin, and the Coast 

 and Geodetic Survey's determination of that city furnishes the 

 initial astronomical location. The triangulation in Arizona, with 

 that of a considerable area adjacent to it in New Mexico, Utah, and 

 Nevada, starts from a base near Fort Wingate, N. Mex., and rests 

 upon the astronomic determination of that place. The triangulation 

 in Oregon and California rests upon lines furnished by the Coast and 

 Geodetic Survey ; while that in Montana rests upon a base measured 

 near Bozeman, and upon the astronomical determination of that 

 place made by the Wheeler Survey. The topographic work of this 

 section is done by plane-table, supplemented in greater or less de- 

 gree by traverses. The plane-table work is regarded, however, as 

 a sketch, and, coincidently with it, a secondary triangulation is 

 carried on with theodolites, which, when platted, senses to correct 

 the plane-table sketch. Heights are measured with the barometer 

 and the vertical circle. 



The quality of the work is to be measured first by the accuracy 

 of the geometric control ; second, by its quantity, i.e., by the num- 

 ber of located points per square inch of map-surface ; third, by the 

 distribution of these points ; and, fourth, by the quality of the 

 sketching, by which the geometric skeleton is filled out into the 

 proportions of the map. The angles in the primary triangulation 

 are read by instruments having circles 6 to 1 1 inches in diameter, 

 reading to 5 or 10 seconds. The mean closure errors in the various 

 sections are as follows : — 



relating to the work of 1886. This shows the number in each 

 square inch of map of occupied points, of points located by angles, 

 of inches of traverse line, and of traverse stations : — 



Appalachian. 

 Kansas 



Arizona 



California.... 



Montana 



•6.60 

 6.13 

 9.05 



In the above table the points located by the two methods of 

 intersection and traverse are given separately. This has been done 

 because they differ in value. Those made by intersection are 



Within this primary work, a secondary system is usually carried 

 on, with minute-reading theodolites ; and, resting upon these loca- 

 tions, large numbers of minor points are determined by the plane- 

 table, or by traverse, coincidently with the sketching of topography- 

 Thus as the lines to be determined become shorter, and the proba- 

 bility of an accumulation of error less, the means provided for their 

 measurement are proportionately less accurate, until, in the ulti- 

 mate work, — that of sketching, — the only means of measurement 

 are the eye and hand of the topographer. It does not follow from 

 this, however, that any part of the geometric work is in appreciable 

 error. It is required that all location shall be sensibly accurate 

 7tpcin the map, and this condition is everywhere fulfilled. 



Second, the number of located points, or the amount of geomet- 

 ric control, varies with the character of the country. The number rri„,-^m 

 of such points is necessarily greater in a country of high relief than 

 in one of low, rolling hills ; it is greater in a country of small, ab- 

 rupt features than in one of large features ; it is greater in a well- 

 settled country, containing many cultural features, than in an un- 

 settled one. Consequently, in this regard, the work done in differ- 

 ent areas varies greatly, as will be seen by the following table. 



'. .'^ \ ' .""* 



.. . J* . • , 



selected points, chosen because of their value in controlling area, 

 while of the traverse locations a large proportion have no value 

 whatever except for the purpose of carrying forward the line, and 



