ANTS VS. COTTON-WORMS. 181 



net proceeded to deprive the fly of its legs and wings, which -were allowed to fall to 

 the ground. The fly was then carried away. Under these same trees I found the 

 wings of Alctia moths, and it looks from these as though these moths are sometimes 

 killed by the hornet ; still, I never saw a hornet in the act of killing a moth, or with, 

 the dead body of one, and I am aware their usual food is flies. 



We find, then, that certain species of wasps destroy the cotton-worm, 

 and also, without much doubt, the cotton-moth. The following species 

 of so-called "wasps" were caught on the cotton plant in Alabama, and, 

 in all probability feed upon the worms :* Monedula Carolina, Fab. (Hy- 

 menopt., fam. Bembecidae; Ells k-notata, Fabr.; Elis plumipes, Drury 

 (Hymenopt, fam. ScoUadae); Pelopceits cccruleun, Linn. (fam. Spliegldae}-, 

 Polistes bellicosa, Cress, j Vespa Carolina, Drury. 



ANTS (Hymenopt., fam. Formicariae). The predaceous insects from 

 which the cotton-worm suffers the most are, without doubt, the ants. 

 These insects, from their war-like habits and the enormous numbers in 

 which they occur, seem peculiarly fitted to hold in check even so dan- 

 gerous an enemy as the cotton-worm. The efficacy of ants as cotton- 

 worm destroyers has been noticed by but few writers upon the cotton- 

 worm, and indeed there are some who insist that they never attack it. 

 During my own stay at the South I never was able to see ants attack a 

 worm upon the plant. Upon the ground, however, the case was far dif- 

 ferent, as is shown by the following brief extracts from my note-book: 



August 28, 1878, 8 a. m. I revisited the field ; there are many larvae crawling over 

 the ground. I have collected specimens of a small ant, which I find destroying these 

 larvae. The head and thorax aie brown, while the abdomen is shining black. They 

 sting severely. 



A perfectly healthy cotton-worm is crawling along the ground, an ant rushes up to 

 it, and, I presume, stings it; the larva at once wriggles away a short distance an 

 inch or so (as the larva cannot get a firm hold on the ground, it is unable to spring 

 as when upon a leaf) ; the ant follows and repeats the attack. I have seen these ma- 

 neuvers repeated many times. It often happens that the larva escapes, but fre- 

 quently it is overpowered by many ants and destroyed. 



I saw a larva wriggling; a single ant was clinging to it, and, although the larva 

 struggled violently, the ant kept its hold. Soon other ants sprang upon the larva, 

 and within two minutes it was overpowered. This occurred over a crack in the 

 ground from which the ants emerged. 



In dry weather the ground cracks to a great extent. The ants make their nest in 

 these ci acks, and while excavating them cover the sni face of the ground with fine 

 particles of earth. It is difficult for cotton-worms to crawl over such places; for 

 when they seize hold of the loose particles of earth by their fore legs, they are unable 

 to balance themselves, roll over upon their sides, and if the earth be hot, speedily per- 

 ish. In this indirect way the ants cause the destruction of millions of the worms. 



I sent from Bacouton, Ga., spe/;hn<>ns of an ant which I found there attacking and 

 destroying Alctia larvae which were crawling on the ground. This ant does not seem 

 to have the power of stinging, but worres the larva to death by biting. 



The first notice of the services of ants as cotton- worm destroyers that 

 we have seen was by Mr. Winfree, in De Bow's Review for 1847. He 

 stated that the ants were a wonderful check to the multiplication of the 

 * Identified by Mr. Cresson. 



