334 THE NEW WORLD OF SCIENCE 



NATIONAL ARMY Continued 



1 Gordon Atlanta, Ga Alabama, Georgia, Tennessee 



Grant Rockford, 111 Illinois, Wisconsin 



Jackson Columbians. C. .... Florida, North Carolina, South 



Carolina 



Lee Petersburg, Va Pennsylvania, Virginia, West 



Virginia 



Lewis American Lake, Wash. Alaska, California, Idaho, 



Montana, Nevada, Oregon, 

 Utah, Washington, Wyoming 



Meade Annapolis June., Md.. District of Columbia, Mary- 

 land, Pennsylvania 



Pike Little Rock, Ark. ... Alabama, Arkansas, Louisiana, 



Mississippi 



Sherman Chillicothe, Ohio .... Ohio 



Taylor Louisville, Ky Illinois, Indiana, Kentucky 



Travis F. Sam Houston, Tex. Oklahoma, Texas 



Upton Yaphank, L. I., N. Y. New York 



NOTE: The States indicated represent the chief source of troops at 

 each place. There are small increments from other points in a number 

 of camps. 



1 This camp was occupied by the troops indicated less than two 

 months, when these troops were sent to Wheeler and replaced at Gordon 

 by draft men from many States. 



By comparing Tables 2 and 3 it may be seen that climate 

 had no recognizable influence on the death rate. Take a map 

 of the United States and locate the camps. It will be found 

 that those with high, intermediate and low death rates enjoy 

 much the same climate ; for example, \Vheeler, rate 28.3 ; Jack- 

 son, 19.9; Sevier, 15.5; Hancock, 2.6 lie in the same region. 

 Now, take another map and locate the camps not where they 

 are but in the region from which their troops come. Then, it 

 will appear that that portion of the United States east of the 

 Mississippi, and north of the Ohio and Potomac rivers contains 

 no camp with an annual death rate of 8 per 1000. Moreover, 

 every camp with similar low death rate, with two exceptions, 

 lies in this area. One of the apparent exceptions, Camp Gor- 



