HIGHWAYS OF DISPERSAL. 43 



ally distributed ; whereas along the Atlantic Coastal Plain its movements 

 were irregular and uncertain, and although its advance still continues, com- 

 plete occupation of the region has not thus far been effected. 



In 1872 and 1873 there branched off from the general horde advancing 

 eastward through Illinois, Indiana, and Ohio several columns which moved 

 southward along the river valleys of the Allegheny Plateau, and in the years 

 following showed a well-marked trend of migration. About 1891 or 1892 

 these columns, having worked slowly down the western side of the Appa- 

 lachian barrier and extended southward beyond it, united with the southward 

 trend of migration along the Piedmont belt, thus completely encircling the 

 Appalachian Mountains. These two united columns have since 1894 or 1895 

 been working slowly southward, but have not as yet been able to reach the 

 Gulf coast. 



The topographic area comprised within the Ozark Plateau and the Ouachita 

 Mountains, situated between the Great Plains on the west, the prairies on the 

 east, and the Coastal Plains area on the south, is one well adapted to retard 

 migration, if not to serve as an effectual barrier. Unfortunately, almost 

 nothing is known concerning the dissemination of L. decemlineata in this 

 region. We know that from middle Missouri it extended very slowly to the 

 southern counties (Riley), but this may well have been due to other than 

 topographic causes. Over much of this area it is not common, and it is 

 probable that in many places it does not exist. A possible trend is indicated 

 to the east of the Ozark Plateau, which passes the eastern end of the Ouachita 

 Mountains through Arkansas, into northeastern Texas, where decemlineata 

 is found. This trend arose from the Mississippi Valley movement and may 

 be considered to be a part of it. 



It may be of value at this point to compare the dissemination of L. decem- 

 lineata with that of other forms that have spread over the same regions and 

 have been influenced by the same natural causes. The data of the dissemi- 

 nation of Pieris rapce L. has been gathered and published by Scudder (1S87), 

 and although he does not place any stress upon the relation of the spread of 

 this insect to topographic and climatic conditions, a study of the published 

 data and lines of yearly advance shows the same general control of dissemi- 

 nation as that found in L. decemlineata. It is of interest to compare the 

 spread of these two insects, one a native of the Great Plains, which began its 

 advance eastward at about the same time that the other, a native of Europe, 

 was introduced into the east and began its march westward. Although the 

 advance of Pieris rapce has been against the action of powerful climatic factors, 

 the lines of its most rapid and successful advance have been along such natural 

 highways as those afforded by the St. Lawrence-Great Lakes system. Other 

 species of insects have responded in like manner to these same influences. 

 Webster, in his study of the distribution of the chinch bug, has come 

 to the conclusion that it entered the United States from the south through 



