82 



Instead it now works upward, forming- a curved tunnel, with smooth 

 walls of well-packed earth webbed together by a thin coating of silk. 

 (See PI. XIII, fig. 1.) Normally the tunnel stops abruptly about one- 

 eighth or one-fourth of an inch below the surface, leaving a thin wall 

 of earth through which the moth must penetrate on emerging. As 

 soon as the burrow has been completed, the larva repairs to the lower 

 end (PI. XIII, fig. 1) and there transforms into the pupa, which lies in 

 the cell, with the head slanting upward, until the emergence of the 

 moth. A number of plaster casts of pupal burrows were made, and 

 a few of them which show the typical variations in shape and size are 

 figured on the accompanying plate (PL XIII, fig. 2). 



VARIATIONS IN THE FORM OF THE PUPAL CELL. 



There is a very evident relation between the moisture content and 

 character of the soil and the form of the pupal cell. In dry soil of a 

 sandy nature the larvae always dig deeper than in moist soil of the 

 same composition. (See fig'. 8.) In the ''black-waxy" soil of north 

 Texas ver^^ shallow burrows were made, due no doubt to the mechan- 

 ical difiiculty offered by its tough consistency. In some cases the larvae 

 could make no impression on dry, baked soil of this kind and were 

 compelled to crawl into the cracks caused by drying. 



In the accompanying table are given the dimensions of eight pupal 

 cells made in moderately moist sandy loam. They were obtained by 

 making plaster casts of the cells. These show a considerable individual 

 variation. They average 46 mm. in depth and 43 mm. in horizontal 

 extent. 



Table XXXI. — Dimensions of pupal cells. 



Vertical 



Horizontal 



extent. 



extent. 



Millimeters. 



Millimeters. 



53 



48 



71 



57 



80 



50 



38 



40 



26 



43 



20 



39 



57 



27 



20 



39 



Temperature also exerts an influence on the depth of the burrow. 

 LarviB of the late fall brood dig much deeper than those pupating 

 earlier in the season. In a lot of seven overwintered pupse, at Hetty, 

 Tex., the depth of the respective cells in millimeters was as follows: 

 50, 50, 125, 125, 125, 150, and 175, with an average of 114 mm. At 

 Paris, Tex., six cells averaged 80 mm. in depth in black cla}^ loam, 

 and 13 in sandy loam averaged 98 mm., as compared with 46 mm. for 

 the August brood. The reason for this .is evident, since the over- 

 wintering pupae must withstand many adverse conditions to which the 

 summer broods are not exposed. 



