RADIATION AND TEMPERATURE MEASUREMENTS. 31 
heat radiation. However, it is not considered a true endothermic phenome- 
non caused by the absorption of heat by the photogenic organs, but is rather 
a slow conduction of heat from the walls of the radiometer to the insect, 
which is at alower temperature. If the phenomenon is one of a true trans- 
fer of radiant energy, then the galvanometer should have reached its maxi- 
mum deflection in 4 seconds instead of 60 to 80 seconds, and it should have 
remained at a maximum instead of slowly drifting back after the expiration 
of 1.5 to 2 minutes. The crucial test would be to use a rock-salt window 
which would permit the long-wave radiation from the thermo-junction to 
pass directly to the insect, instead of impinging on the glass walls of the 
radiometer. ‘This cooling effect led to the thermo-couple explorations to be 
described presently. 
The radiometric tests of the flash and the glow of both Photuris pennsyl- 
vanica and Photinus pyralis extend over a period of 3 weeks, during which 
time four distinct series of measurements were made; each series took from 
1 to 3 hours, during which time various experiments were tried, using red 
absorption glass, etc. In the earlier measurements on the Photuris the 
insect was attached to a thin wooden rod about 15 cm. long, held horizon- 
tally and suspended over the thermo-junction. ‘The tests were made in a 
darkened room and shields were provided so that no stray light could get 
into the thermopile case. The flashes were as frequent as 3 times per 
second, but in neither the flash nor the steady glow could any radiation be 
detected with certainty other than the cooling effect (conduction of heat) 
which will now be discussed. 
If it were not for the novelty of the phenomenon and the averseness with 
which one views data of this type, after having been taught to search for a 
heating effect rather than the opposite, it would be possible to urge that the 
emission of light is accompanied by an absorption of heat. For example, a 
strong, active Photuris, flashing 80 to 90 times per minute, caused a cooling 
effect of 5 to 6 cm. deflection (galvanometer period 4 seconds, therefore very 
steady and no ‘‘drift); and later on, when the flashes became less frequent, 
in two distinct series of measurements the maximum (cooling) deflection 
was in each case only 3.5 cm. Another specimen gave a (cooling) deflection 
of 1.5 cm. and then began ‘‘flashing brighter”’ (as the records have it), when 
the deflection increased to 3.1 cm. Still another specimen of Photuris, 
‘flashing 3 times per second,” gave a maximum (cooling) deflection of 6 to 7 
cm., and finally only 2.2 cm. deflection. The records show several similar 
cases. Whether these measurements have any significance it is difficult to 
say. ‘The closer to the radiometer the greater was the cooling, but it is not 
known whether the insect was accurately at the same distance from the walls 
of the radiometer when flashing and when it became less active. 
If the insect generates only the radiations which are visible to the eye, the 
heating value of these rays will be small. It is evident that but little infra- 
red radiation can be generated, for so much would be absorbed that unless 
it has some cooling device the temperature would rise to a high value and 
the insect would be speedily desiccated. It is therefore possible to have the 
radiation (heating) effects on the radiometer masked by the thermal con- 
duction effects (cooling) if the temperature of the insect is less than that of 
the thermopile. 
