1902.] NATURAL SCIENCES OF PHILADELPHIA. 615 



marked N, and a score or two of Petri cells about teu centimeters 

 broad and one centimeter deep. All these habitations contained 

 flakes of sponge saturated with clean water, and were suitably 

 supplied with morsels of food for the ants. In none of their resi- 

 dences did I ever disturb the ants oftener than twice in a single 

 day by any alteration of their environment. 



I put papers of vivid colors under the transparent glass floors of 

 these habitations, offering always a choice between colors. In the 

 maze four colors were presented; in the nest, three; in the Petri 

 •cells, two. During two months of experimentation iu this line 

 the ants never moved their young, nor themselves congregated 

 upon one color in preference to another, nor did they in conse- 

 quence of a change in the color of their floor, or of an interchange 

 between black and white, ever remove from any spot upon which 

 they had congregated. They gave no sign of regarding any 

 quality in their floor other than its opacity, and the most vivid 

 reds, greens and blues displayed there in full daylight failed either 

 to entice or to agitate them. 



Their behavior was very different when colored glass was used 

 for transmitting the light that entered their abodes. In all cases 

 diffused, not direct, sunlight was used in these experiments. The 

 walls as well as the roof of the dwelling Avere transparent, and the 

 superimposed colored panes extended two inches beyond the walls, 

 so that the chosen light entered from every quarter except through 

 the floor, which Avas ahvays opaque white. This made the condi- 

 tions nearly equal throughout the abode. The dwellings were 

 placed upon a large table where the summer temperature was the 

 same on all sides, and the diffused light from a large north window 

 gave fairly equal illumination over the whole.* 



The ants were not affected by the intensity of the illumination from 

 rays of longer wave-length than violet. — I repeatedly interchanged 

 the panes over a Petri cell, one-half of which was covered by a 

 pane that shut out all rays except a few red ones, and the other 

 half by a pane transmitting much red light. During a residence 

 of several weeks in this cell the ants showed no preference for the 

 less intense light. 



I also covered one-half of a cell with a colorless pane, and the 



* These experiments were made at the Marine Biological Laboratory 

 At Wood's Hole, Mass., in June, July, August and September, 1903. 



