SYNOPSIS OF ANSWERS TO QUESTIONS ON WIND. 123 



migration, therefore, as affected by winds, is narrowed down to these 

 two considerations : 



1. Has a south or southerly breeze, with a velocity of 12 to 20 miles 

 an hour, sufficient force to bring the cotton- worm moth from the "peren- 

 nial cotton-fields ? " 



2. Would an occasional strong breeze of one, three, or five days' dura- 

 tion, at the rate of velocity given, allow a sufficient length of time for 

 the moth to make its journey ? 



If these questions can be answered affirmatively, having in mind that 

 the moths are moved by the winds within our own borders, we have a 

 very strong argument in favor of the theory that the insects are brought 

 to the cotton States from the southward at the beginning of the season, 

 for the winds are almost constantly from the southward through the 

 season when the moths would most likely take advantage of them. The 

 ordinary low rate of velocity (for night winds) often increases to 12, 15 r 

 18, or 22 miles per hour, and this force, with slight variation, occasion- 

 ally continues for two or three, or even five days in succession. If, on 

 the contrary, a higher velocity is required than the figures in the tables 

 indicate, and for a longer duration of time than has been stated, the wind 

 records prove that the destruction of 1873 was not occasioned by the 

 progeny of moths that came across the Gulf of Mexico. 



In the circular sent out by the Department of Agriculture, the corre- 

 spondents were requested to furnish data upon the two points named 

 prevailing direction and velocity for each month separately from Feb- 

 ruary to June, and collectively from July until frost. Keplies were re- 

 ceived from over seventy localities in the cotton States, the greater ma- 

 jority from Georgia, Alabama, Mississippi, and Texas. As a whole r 

 these observations bear evidence of reliability, though in some cases the 

 question has been misunderstood, and two or three points of the compass 

 indicated at once as the "prevailing" direction for the month. In view 

 of this error in making up the returns it was found necessary, for the 

 sake of approximate correctness, to tabulate all of the observations, and 

 thus ascertain the prevailing direction by calculating the percentages 

 for each point of the compass. The figures thus obtained cannot, of 

 course, be relied upon as absolutely correct, as the replies are not given 

 from actual records, or for a particular year, but as general observa- 

 tions. The reply of " variable," which occurs occasionally in the returns, 

 necessarily alters the final figures in estimating the percentages of winds 

 from the different points ; still the results obtained agree very closely 

 with the official figures, as ascertained from the records of the United 

 States Signal Office. 



As has been stated, the replies are principally from four of the Gulf 

 States, though ten States are represented in the returns. The total num- 

 ber of answers for the different months amount to about four hundred. 

 These observations are for day and not night winds. For convenience of 

 tabulation, as well as to save unnecessary labor, but four " directions" 



