52 ROBERT STANLEY McEWEN 
ments, and while it is possible that the desire for a certain 
outcome might cause an unconscious variation at times, I think 
it very unlikely that this was of sufficient magnitude to seriously 
alter the results. 
For the purpose of providing a constant standard source of 
light a 200 watt nitrogen filled Mazda lamp was suspended within 
a tin box 24 em. high by 14 em. wide and 14 deep. Small holes 
were made in the bottom and the top of this box to allow for 
ventilation, while in its front, just opposite to the concentrated 
filament of the lamp, was cut an aperture 3.5 cm. in diameter. 
In front of this openingjwas dastened a flat flask containing water. 
A second box of the same dimensions was now fitted to the 
front of the first and in its front a-hole was cut equal in size to the 
hole in box number one and exactly on a line with it. This aper- 
ture was fitted with clamps for holding the flasks which were to 
contain colored liquids when such were desired. If only white 
light was wanted the outer as well as the inner flask contained 
clear water. This arrangement of two flasks not only shielded 
the insects from heat rays, but also prevented the colored liquids 
from becoming hot. The flasks used were about 1.5 em. thick, 
and identical flasks were employed in the same positions through- 
out the experiments. 
This apparatus was set up at the end of the table on which 
the testing tube was fastened, in such a manner that the center 
of the lamp filament was exactly opposite to the end of the tube 
and about 24 em. away from it. This made the end of the tube 
about 3 cm. from the outer flask. Finally a blackened screen 
was set up between the end of the tube and the light boxes, with 
an aperture in it opposite to the end of the tube and about the 
size of the flasks. The purpose of the screen was to shade the 
tube from the slightly diffused light which escaped from cracks 
in the boxes and through the holes above the lamp. The room 
where most of the experiments were performed had its walls 
painted a dead black, and the windows were provided with black 
opaque shades. All timing was done with a good watch placed 
in such a position that the face was illuminated by light issuing 
from the aperture. A diagram of this apparatus is shown in 
figure 1. 
