LEPIDOPTEEOUS PUP^E AND TIIEIE SURROUNDING SURFACES. 
355 
Of the 12 crowded pupae, 7 were suspended from the roof and were all 
very light.(3), very uniform. 
„ ,, 3 were suspended from the side against the 
white background, 
of which 1 was suspended from a point 3 
mm. below roof; 
1 was suspended from a point 2'5 
centimetres below roof; 
1 was suspended from a point 5’0 
centimetres below roof: 
Of the 12 crowded pupae, two were on the white opal floor and not fixed to anything. They were 
both (4), typical. 
The relative positions of these 12 larvte are shown in Fig. 2. Natural size. 
Fig. 2. 
Very light (3), very uniform, and not 
much gold. 
1-7 indicate the positions of the 7 pupae suspended from the roof. 
a. A pupa was attached to the side of cylinder 3 mm. below this point. 
b. „ ,, ,, 2‘5 centimetres below this point. 
™ ,, ?) 5 0 ,, ,, 
X X The two pupae on the floor were probably suspended from these points in the larval state. 
The inner circle represents the actual size of the roof, looked at from below. 
x. The space between the two concentric circles represents the thickness of the glass sides 
of the cylinder. 
The sides of the cylinder which bound the area below A—B were clear and uncovered 
by paper, and turned towards the light. 
The fine line outside the cylinder above A—B represents the paper background. 
Nothing could we]l be more conclusive than the above comparison ; of the crowded 
pupse, 9 out of 12 come within a very short distance of each other, and are all (3) ; 
while the remaining three are more isolated in position, and the two most isolated of 
these are (4). Furthermore, those more completely isolated in separate cylinders are 
(4) and (5), not one of them reaching (3). The experiment (with others already 
described) seemed so conclusive that I did not consider it necessary to seek further 
2 z 2 
