18 THE SOLITARY WASPS. 



This is wifrh-out the slig-htest stimulation as we do not even touch the 

 box nor the shelf upon which it stands. The movement is not at all 

 violent as in other caterpillars (see Nos. 15 and 56). The muscles have 

 relaxed a little so as to let down the anterior segments, but the body is 

 still in a curled up position. 



September 5. The caterpillar is alive and is still tightly curled up. 



September 6. It now lies flat. The body is shrunken and the colo-r 

 lias faded to a li\-id blue. 



September 15. The caterpillar looks as if it were dead but still re- 

 isponds to careful stimulation. 



September 23. The caterpillar is now of a sickly yellowish hue and 

 is shrunken to a quarter of its original size. We get a scarcely per- 

 ceptible response to stimulation. 



September 25. The caterpillar is unquestionably dead. 



Among tlie fifteen caterpillars that we have taken from the 

 nests of nrnaria three kinds are represented, twelve of them 

 belonging to one species, two to the second, and one to the third. 



The egg, which is laid upon the side of the sixth or seventh 

 segment* (PL VIII., fig. 5), hatches in from two to three days; 

 the larva spends from six davs to two weeks in eating, and then 

 spins its pale yellowish cocoon. 



The nesting habits of urnaria closely resemble those of the 

 other members of the genus, as reported by various observers. 

 The spot chosen is in firm soil, sometimes in open ground but 

 much more freqeuntly under the leaves of some plant. The 

 plan is a very simple one. A tunnel of about an inch in length 

 leads to the pocket in which the caterpillars are stored. There is 

 no hardening of the walls in any part. We took pains to draw 

 every nest that we opened, and, as will be seen from the illus- 

 trations of some of them, there was a very considerable variation 

 in the minor details, such as the obliquity of the entrance tun- 

 nel, the shape of the pocket, and the angle at which the tunnel 

 and pocket were joined. . (PL VIIL, figs. 1-4.) 



The work is done with the mandibles and the first legs. 

 When it has proceeded so far that the wasp is partly hidden, she 



*In the drawing the egg is, by mistake, shown upon the eighth seg- 

 ment. 



