29 



of an inch from the repellent. The experiment was closed by the re- 

 moval of the sand, the odor of which was scarcely distinguishable. 



Experiment 5777 , Oil of Lemon. — May 27, 2: 37 p. m. This ex- 

 periment was made in a cage with a 3-inch cement disk at the center of 

 the base, and a circular orange cover 3^4 inches across. Sand mois- 

 tened with oil of lemon was placed beneath the center of the cover. 

 The ants were thrown immediately into great confusion, and about half 

 of them at first ran from under the cover, deserting for a time the 

 larvs nearest the sand. These were about one and a fourth inches from 

 the repellent. In 5 minutes the ants on the base were beginning to re- 

 turn beneath the orange cover, some moving the larvae farther from 

 the sand and recovering those at first deserted. After 10 minutes the 

 distance from the sand to the nearest larva was five-eighths of an inch. 

 They were still moving their larvae back. 3 : 08, nearest larva now 

 six-eighths of an inch. 3 : 40 (63 minutes ), nearest larva seven-eighths 

 of an inch. Most of the ants are out on the base (Fig. 7, A). 5 and 



Fig. 7 



5:30, situation unchanged. May 28, 8 a. m. (17 hours, 23 minutes), 

 larvae are more scattered, and the inner edge of the pile is about five- 

 eighths of an inch from the sand. About three-fourths of the ants 

 are under the cover with them. 2 and 5 : 30 p. m., nearest larvae six- 

 eighths of an inch from the repellent. May 29, 8 a. m. (41 hours, 23 

 minutes), nearly all the ants are under the cover. The nearest are 

 three-eighths of an inch from the repellent, as shown by Figure 7, B. 

 11 : 50 a. m. and 1 : 30 p. m., situation unchanged. 5 : 45, ants now all 



