618 



PROCEEDINGS OF THE ACADEM'X OF 



[Sept. 



N. 



Experiment 2. — Using the artificial uest represented in the draw- 

 ing, N", I put in, on June 9, five queens, thirty workers, and a 

 teaspoonful of larvse, all taken from a wild nest on the preceding 

 day. The food-roora was roofed with colorless glass, was used by 

 the ants solely as a feeding place, and does not otherwise enter into 

 the experiment. Over each of the three other rooms I placed a 

 pane of colored glass — 



Red, transmitting red rays only. 



Yellow, transmitting red, orange, yellow, green, blue, indigo. 



Blue, transmitting orange, yellow, green, blue, indigo, violet. , 



la 14 trials, occupying 14 days, the ants always removed from 

 under the blue pane, whenever it was placed over them, and took 

 shelter in another room. They removed 8 times to red, and 6 

 times to yellow. 



In this experiment they always departed from the room to which 

 violet rays were admitted, and settled in one of the two other 

 rooms. 



In order to secure an expression from the ants concerning partic- 

 ular colors I used Petri cells, offering a ch lice between only two 

 panes. I made each test with newly captured ants, including one 

 or more queens, ten or more workers, and some inert young. 



Experiment 3. — Time, 40 days. 



Red pane, transmitting red rays only. 



Treble orange pane, transmitting red, green. 



The ants showed no preference for either side of the cell. ' The 



