RADIATION AND TEMPERATURE MEASUREMENTS. 35 



and more especially because of the question of the reality of the results. 

 The tests were repeated when doubts arose at various times during an inter- 

 val of more than 3 weeks, the instruments having been dismounted in the 

 meantime, and they always led to the same result, viz, that the temperature 

 of the insect is less than the air and that the region of the luminous segments 

 is hotter than the dark segments. 



In these experiments no attempt was made to obtain exact measurements 

 of the difference in temperature of the luminous and the non-luminous seg- 

 ments. Indeed such data would be illusory. A rough estimate of the 

 temperature may be obtained from a knowledge of the galvanometer sensi- 

 tivity, the thermo-electric power of the junctions, and the resistance of the 

 circuit. Using these constants, and assuming a deflection of 20 cm., the 

 equivalent difference in temperature is 0.6°. Frequently the indicated 

 temperature was only 0.3° to 0.5°, while the "off the scale" deflections indi- 

 cated temperature differences as great as 0.8°. The bright luminous points, 

 L', Fig. I, on the female pyralis indicated a temperature of 0.2° above the 

 surrounding dark regions. With the thermo-couples it was not possible to 

 determine with certainty whether the deflections were increased (a heating) 

 during the flash. The experiment showed that this temperature difference 

 was present continuously in healthy specimens. This would seem to indi- 

 cate that even if the temperature difference is due to some biochemical 

 reaction, the light emission is under the control of the insect. The breathing 

 spiracles being over all of the segments, it is difficult to understand how the 

 influx of oxygen could produce a local effect in the luminous organs, unless 

 there is some chemical action in these organs. 



The galvanometer deflections were of the same magnitude for the cooling 

 effect as they were for the heating effect. This seemed to indicate that the 

 temperature of the insect was not much below that of the room (26° C.) at 

 the time of the experiment. Hence, having been informed that the temper- 

 ature of these insects is about 17° to 18°, and that these measurements had 

 been made by placing a number of insects in a vessel containing a thermom- 

 eter, the writer undertook a similar experiment. Two similar glass speci- 

 men tubes, 15 cm. long, 2 cm. diameter, and two similar thermometers were 

 provided. The thermometers passed through corks fitted into the tubes. 

 These tubes were placed side by side in a vessel containing cotton batting. 

 More than 100 Photinus pyralis were caught and shaken directly from the 

 net into one of the glass tubes, which was then about half full of insects. 

 After replacing the tube in the insulating material and waiting 10 minutes, 

 the writer was surprised to find that the temperature had risen 1° above 

 the empty tube instead of dropping to a lower temperature. Table 4 shows 

 the course of events during the test. During the evening, as the activity 

 of the insects decreased, the temperature difference decreased. Inter- 

 changing the places of the test tubes showed no variation in the readings, 

 and the conclusion arrived at was that the temperature rise was due entirely 

 to the presence of the insects. In view of the fact that the body temper- 

 ature of these insects appeared to be almost as much below the room as the 

 temperature of the luminous segments was above that of the insect, it is not 

 unthinkable that in the thermally insulated vessel it was possible for the 

 temperature to rise as observed. 



