PHOTIC BEACTIONS OF HONEY-BEE 383 



The objection might be raised that although the majority of 

 values of d are positive, a number of them are too small to be of 

 any significance. It is true that differences of the order of l°/cm. 

 or less might easily be attributed to slight errors in tracing the 

 course of a bee. Errors in recording, however, are as likely to 

 occur in one direction as the other. Such is not the case with 

 these small values. The class of d values betwen and — 2°/cm. 

 contains but fifteen, while the class of to +2°/cm. numbers 

 twenty-nine — nearly twice as many. 



Moreover, the mode of the curve, -t-2°/cm. to -[-4°/cm., lies 

 well beyond these small values. Differences of this magnitude 

 are readily detected in the records of bees which circled constantly 

 toward the functional eye, as the pair of determinations in figure 

 12 demonstrate. Each record shown in the figure was of exactly 

 thirty seconds' duration. The first two were taken three min- 

 utes apart in 24 mc. illumination, while the third and fourth 

 were taken about twenty minutes later in 957 mc. illumination, 

 one minute apart. The value of d in this case is -|-2.98°/cm. — 

 about the average modal value. 



In bees exhibiting considerable variation in their deflections, 

 however, d values, or of the modal class of even greater magni- 

 tude, are not so easily recognized without the accompanying fig- 

 vires. To illustrate this, I have selected a pair of determinations 

 (fig. 13) approximating the mean value of the frequency polygon, 

 which is +4.35°/cm. As attested by the data given in connec- 

 tion with the figure, this animal was extremely variable in its 

 direction of turning. All four records were taken in the brief 

 period of eight minutes, nos. 1 and 2 being of twenty-eight sec- 

 onds' duration each; nos. 3 and 4, of thirty seconds. The value 

 of d is +4.09°/cm. — a fact which does not become apparent until 

 the records are submitted to a careful scrutiny. 



The data presented in table 2 and figure 11 show clearly that 

 bees with one eye blackened tend to turn more toward the func- 

 tional eye in a non-directive illumination of 957 mc. than they 

 do in one of 24 mc. The validity of this conclusion is confirmed 

 by still another line of evidence. A certain number of animals 

 were found to exhibit a very constant tendency to turn toward 



