Wasps, Bees, and Ants 551 



the far end are dry. He gives the ants no soil, forcing them to use the already 

 made chambers. This formicarium reveals, therefore, none of the secrets 

 of nest-building, but it does reveal admirably a host of those interesting pro- 

 cesses connected part cularly with the life-history of the individuals of the 

 colony. Miss Fielde uses still another kind of nest, also like Janet's with 



FIG. 753. A Janet nest in vertical section, w.c., wet chamber; i, 2, 3, brood-chambers; 

 o., circular openings for brood-chambers made in c., a transparent cover; o.c., glass 

 cover in three removable pieces; d.p., opaque cover; b.p., base plate. (After Janet.) 



fixed chambers, but made wholly of glass, the requisite moisture being fur- 

 nished by a bit of sponge kept soaked with water and placed in one of the 

 communicating chambers. Fig. 754 with its caption explains the make-up 

 of a Fielde nest. 



In the study of the life of ants by means of such formicaries as have just 

 been described, as well as through observations in the field, the student, 

 amateur or professional, should keep in mind certain particular desiderata 

 in formicology. It is highly de- 

 sirable to determine for as many 

 species as possible the exact 

 method of founding a new colony: 

 isolate a queen in a small artifi- 

 cial formicary, well provided 

 with food, and see if she can and 

 will begin one; isolate a small 

 group of workers with some eggs 

 or young larvae, but without a 

 queen, and see if they can and do 

 produce a queen and establish p IG . 7S4 . plan of the Fielde ant-nest, 10 

 themselves as a permanent com- inches by 6 inches, a, entrance and exit to 

 munitv The characteristic hahiK food-rooms (i); 2, nursery; 3, sponge-room; 



b, screens; m, passage, 

 of feeding the young should be 



determined for various species; the presence of or possibility of producing 

 ergatoid (wingless, worker-like) fertile females and males in the case of vari- 

 ous species should be noted; and special attention should be given in all 

 observations to determining in how far the behavior in general, and single pro- 

 cesses in particular, can be explained as machine-like reflexes of unintelligent 



