Honeybees two or three at a time, were placed under a large petri dish into which entered two narrow opposed beams of light, the intensity of one of which could be varied by means of a rotating disc photometer. The bees in trying to get out of the dish ran to the two spots where the beams entered. By counting the number of times they bumped into the glass at each of these two points one could ascertain which light had the stronger stimulating effect. Colors, except yellow, were obtained by the use of Coming glass filters, the yellow filter being a potassium dichromate solution. The transmission curve of each filter was accurately known, the peaks of the curves being well separated. Violet, blue, green, yellow, and red were used. When each of these colors was opposed by a white light it was found that the latter had to be reduced by a certain amount in order to equalize the opposing color in stimulating effect. Using, then, the total radiation per unit area of each filter, as determined at the Bureau of Standards, the relative stimulating efficiency of the colors turned out to be as follows: green 100, blue 54, yellow 29, violet 23, red 2. When these numbers are plotted on the same scale as the curve of the relative sensitivity of the human eye, as worked out by Coblentz, it is found that the curve lies farther toward the violet end of the spectrum than does the curve for the human eye and does not extend as far into the red. It shows decidedly that red does not stimulate bees as much as it does humans, and that violet stimulates them more than it does humans.