THE COTTON-WORM. 145 



walls, and particularly in the recesses of windows, to the great 

 annoyance of the tidy housewife. 



Thus is answered the question, why the cotton-fly did not 

 again eat up the scant foliage which subsequently appeared 

 on the stalks. This little usurper goes forth in search of 

 " whom he may devour," and as soon as he finds a house built 

 and well provisioned, he seizes upon it for his posterity, which 

 he does in the following manner: When he finds a cotton- 

 worm, he pierces it with the instrument with which its tail is 

 armed, and deposits an egg; the cotton-worm soon spins itself 

 up into its case, there to await the period of its perfection, 

 which never arrives, for soon il\e egg of the ichneumon hatches, 

 and falls to devouring his helpless companion. This work of 

 extermination continues until there is not a vestige of the cot- 

 ton-fly left. I venture to say, while I am now writing (1st 

 of December), there is not an egg, chrysalis, or fly, in the con- 

 fines of the United States. My speculations on the nature and 

 habits of the fly have led me to adopt the following hypothesis : 

 That it is a native of tropical climates, and never can pass a 

 single winter beyond them, consequently, never can become 

 naturalized in the United States, or any where else where the 

 cotton plant is not perennial, for nature has made no provision 

 by which they can survive more than ten or twelve days, 

 therefore they must perish wherever the cotton plant perishes 

 during a period of six months. That wherever they have pre- 

 vailed in our cotton-growing regions, it is when they have become 

 very numerous, and consumed all the cotton in their native 

 climes, and then go in search of their food in more northern 

 climates. It is not to be presumed that this happens often, 

 but the same remark will hold in regard to the cotton-fly as it 

 will to many other insects, that owing to some unknown cause, 

 they become exceedingly numerous, but at long and irregular 

 intervals. The locust has already been noticed as an example, 



7 



