GEOGRA PHIC LITER A TURE 231 



is about 18 by 28 inches in size, and is printed in four colors: black f..r 

 projection lines, names, and all cultural features ; blue for streams ; jjrecn 

 for the oceans and large lakes, and brown for the hill and mountain fea- 

 tures. These relief features are shown by contour lines. The contour 

 interval, from 2,000 feet upward, is 1,000 feet. Below the 2,(M)()-foot con- 

 tour the interval is variable. Over this base map the magnetic curves are 

 printed in red. 



The magnetic declination, popularly called variation of the compa.s«, is 

 subject to several known jjeriodic changes. Of these the most importsint 

 is the secular change— a change with a period running through centuries ; 

 hence its name. As this secular change is progressive from year to year 

 for long periods, and as it amounts in the United States to from 2' to '^' 

 per year, it is for the surveyor and mariner the most important of the 

 periodic changes. Indeed, it is the only one of mudi practical imjiortance 

 at present. It is to this }>ractically important (juantity tluit Mr Oannett 

 has wisely devoted the greater part of the labor expended on this memoir. 

 The weakness of similar maps hitherto produced has been recognized l>y 

 both their makers and users to be largely due to defective knowledge of 

 the secular change. 



Of the 237 pages comprised in the memoir 82 are devoted to data for 

 secular change. A table of results by counties occupies 135 pages, while 

 the remaining 20 pages are given to introductory matter, discussion, state- 

 ment of sources of data, etc. 



The sources of the data are the Coast Survey, Lake Survey, the Wheeler, 

 Hayden, and Powell Surveys, New York State Survey, New Jersey Geo- 

 logical Survey, Boundary Surveys, United States Corps of Engineers, .\rmy 

 Exploring Expedition, National Academy of Sciences, and others ; but it is 

 chiefly from the records of the United States General Land Office and 

 from county surveyors that a vast quantity of hitherto unused material 

 has been derived. Indeed, so abundant are data in the General Land 

 Office that it was only needful to select for the older " land office " Stales 

 such as were desired. The mass is much greater than is needed to pro- 

 duce a map sufficient for all practical needs. As to this Mr Gannett says : 

 " I have not attempted to make a complete collection of this material. 

 The amount is too vast to make it worth while. I have, however, col- 

 lected all the observations which appear upon the plats of exterioi-s and 

 standard lines (the Land Office requires that in the survey of all stan.lar*! 

 and exterior lines the declination be observed), supplementing them 

 wherever needed by observations made in connection with the subdivision 

 of townships. Altogether, I have abstracted from the plats of the (ien- 

 eral Land Office nearly 20,000 observations, andtliese form, perhaps, nine- 

 tenths of the material herewith i)resented." 



As the work of subdivision and accompanying magnetic observations 

 began a century ago, it is obvious that these Land Ullice records cnsti- 

 ture a veritable storehouse of information on secular change— a storehouse 

 of which Mr Gannett is the first to make general use. 



In addition to these data a circular was sent to all the county surveyors 

 in the United States, and from the returns much valuable information 

 was obtained. 



