METEROLOGY AND MIGRATIONS. 203 



dieted -J and, again, their arrival in the States east of the great pUiins 

 be announced with a certain degree of precision, as indeed could now 

 in some measure be done from the experience we have already had — but 

 we may be able to foretell the course taken in the return flight of their 

 progeny in the early part of the succeeding season. It remains to de- 

 termine the causes of this return migration, this completion of the " mi- 

 gration cycle," as Mr. Dawson terms it. It is evident that in this case 

 the desire for food is not the cause, for food is many times more abundant 

 in the Mississippi Valley than on the plains whither they return. The 

 solution of the problem, we think, must be sought in the direction of the 

 prevailing winds during the spring months up to the middle of June — 

 the time when they become winged. 



South of latitude 40^ the locusts fly before the first of July from the 

 southeast. In Texas, Missouri, and Kansas, and probably E^ebraska, 

 this is the universal experience, and south of latitude 40^ the winds in 

 the spring are universally southerly ; on the other hand, later in the 

 season, the prevailing winds are northerly or northwesterly. 



North of the fortieth parallel, in Iowa and Minnesota especially, the 

 locusts do not return as a body to the northwest, though many do, and 

 careful observations by experts are needed to determine the relation be- 

 tween the prevailing direction of the winds and the variable and uncer- 

 tain course assumed by the native-born broods of theEocky Mountains. 

 Granting, therefore, this setting-in of southerly and easterly winds, 

 which may last until the locusts are winged, when they rise on the 

 wing into the air they are known to move in a general northwest direc- 

 tion. It is known from the observations made during the past season 

 that they are borne along by these southeasterly winds, and pass 

 over on to the plains from Texas to Colorado, Wyoming, and Ne- 

 braska, as is stated in chapter YIII, to which the reader is referred for 

 a discussion of the immediate or more special causes of migration. 

 The cause is seen, then, to be entirely independent of the question of sub- 

 sistence ; possibly the reproductive instinct; as well as the adverse 

 features of a damp, hot, debilitating climate, like that of Texas and the 

 border States, as compared with the native breeding places of their 

 parents, which cause them to become uneasy, restless, to assemble high 

 jn the air, and seek the dry, hot, elevated plateau of the northwest. 

 Should this be so, the cause of their migrations is probably purely me- 

 chanical. We may expect in the future as in the past that south of lati- 

 tude 40O the spring and early summer migrations from the eastern limits 

 of the locust area will be always toward the northwest, and that the 

 July, August, and early September migrations from the Eocky Mountain 

 plateau will be in a general easterly and southeasterly direction. 



The following remarks on the connection between meteorological phe- 

 nomena and locust migrations have been communicated by Mr. Cleve- 

 land Abbe, who writes as follows : 



It is impossible for me to satisfactorily investigate the question of locust migrations 

 as affected by the weather -without going into a very extensive review of the whole of 



