338 Natural History Bulletin. 



enough to the water's edge to insure continued dampness, and 

 there construct an oval cell of earth, without any admixture 

 of silk so far as I can find. These cells are not simple exca- 

 vations of earth beneath the stone, but are built upon it Hke 

 the cells of some of our mud-wasps and are not very unlike 

 them in shape. A cell in my possession measures 12 mm. in 

 long diameter and 8 mm. across, outside. In it the larva lies 

 curled somewhat in the form of a letter C until changing into 

 a pupa which rests upon its back — a position also assumed by 

 the perfect insect when it is at last disclosed. Escape seems 

 to be made by forcing out one end of the cell as nearly all the 

 empty ones I saw were broken in that fashion. 



The drawings and descriptions are made from full grown 

 larvjE taken from their cells but not from those so far advanc- 

 ed as to be unable to run. All students of Coleoptera know 

 that in the active species the larvae undergo considerable 

 change of form just previous to transformation. 



Tropisternus glaber Herbst. Plate IX, Figs. 13 to 15. 



Color brown above, tubercles and a stripe down the back 

 lighter. Beneath yellowish, the tips of the femora and tibias 

 brown. Head, above brownish with a yellow stripe each side, 

 beneath brownish, with a median and two pairs of lateral 

 stripes yellow, the outer one broader. Prothorax with a 

 crescentic yellow mark on the under surface. 



Form depressed, tapering to both ends. Length 14 mm. 



Head subquadrate, a little narrowed posteriorly, not contin- 

 uing the axis of the body but directed somewhat upward. 

 The upper surface is convex between the eyes but with a flat 

 space near the base; below convex but with a well-marked 

 flattened semicircular area immediately posterior to the men- 

 turn. Frontal margin entire. 



Eyes situated behind the antennae, consisting of six ocelli on 

 each side, arranged in two transverse rows of three each. 

 The outer ocellus of the posterior row is undeveloped in some 

 specimens. 



