77 



Exj^erinient 3. — A newl}^ hatched larva was placed on a plant 3 feet 

 hig-h July 15. It entered the soil July 28, after destroying one 

 square, one flower, and two full-grown bolls. 



Experiment ^. — Larva one-eighth inch long was placed on plant 3 

 feet high Jul}^ 26. It entered the soil August 4, after destroying 

 nineteen squares. No bolls or flowers were injured, although two 

 large bolls bore evident marks where the larva had tasted them and 

 passed on. The daily injury done by this larva may be of interest: 

 First and second da}^, one square; third day, two small squares; fourth 

 day, one small square; fifth da}", one square; sixth day, two squares; 

 seventh day, five squares; eighth day, six squares. 



Experiment 5. — A one-eighth inch larva was placed on a plant 3 feet 

 high July 27. It entered the soil August 7, after destroying four 

 squares, two flowers, and two bolls (one-fourth and three-fourths 

 grown, respectively). 



Experiment 6. — A one-sixteenth inch larva was placed on a plant 3 

 feet high August 10. By August 20 it had destroj^ed thirteen squares 

 and one flower. 



A summary of these six cases shows a total injury of fifty squares, 

 eight bolls, and five flowers. This gives an average of eight squares, 

 one and two-thirds bolls, and one flower for each developing larva. 



A great individual difference in tastes is at once evident, some 

 larvae preferring squares entirely and others feeding more or less on 

 bolls when both are present in sufficient quantities, as they were in 

 all these experiments. Squares are invariably chosen by the youngest 

 larvte, some of them continuing when older to search out the squares 

 and others turning their attention to the bolls. In one case a boll was 

 attacked by a rather young larva, which afterward consumed onl}^ 

 squares and flowers, but this is unusual. On account of their large 

 size, a much smaller number of bolls are destro3^ed than squares. In 

 fact, it was seen in one case that two large bolls and one flower served 

 to mature a larva, while another, which fed only on squares, required 

 nineteen. 



During the experiments each individual plant was covered by a 

 large wire-screen cage, 3 feet square at the base and 4 feet high. This 

 prevented in great measure the escape of larv«, and kept out any 

 other insects which might interfere with the accuracy of the results. 

 The cotton plants each bore from 40 to 60 squares and from 8 to 12 

 well-developed bolls, in addition to a varying number of flowers and 

 small bolls. Hence it is apparent that it would require only seven 

 or eight larvae, feeding at the rate of those observed, to destro}^ nearly 

 all of the fruit on one large plant. 



