586 A new Trap-door Spider. [July, 
on the under side of the partial door, she would carefully move 
it up to the edge. Then placing the distal portion of the palpi 
and anterior pair of legs above, while the proximal portion of 
these limbs and the ends of the mandibles were on the under 
side of the pellet and door, she would fit and press it in shape, as 
one would mold with the hand a moist portion of earth by press- 
ing it into a thin sheet. This is illustrated in Fig. 7, Plate XXIII. 
Indeed it looked very much like the black bony fingers of a hand 
performing the work of pressing. The greatest pressure seemed to 
be brought to bear upon the rounded ends of the mandibles. 
After fastening on a portion thus, the spider would take an in- 
. verted position and apply viscid liquid along the edge and under 
the surface of the door, as shown in Plate xxii, Fig. 6. She would 
then turn about and crawl out for more material. The hole 
being by the side of the jar, I could watch the operation both in 
the hole and upon the cover. By one o’clock in the morning 
(Nov. 13) the door was finished so that the spider could pull 
down the lid, which completely closed the entrance, nicely fitting 
- in around the edge and appearing as if there was no hole nor 
spider, but through the glass the spider could still be seen. 
At intervals during the construction of the door the spider 
would pull it down to observe where the next pellet should be 
placed in order to make the door fit the circular opening of the 
tube. Discovering this she would turn completely around, and 
not being able, with her head in the bottom of the tube, to see 
the place where she intended to put the next load, she would find 
; it by feeling about with her spinnerets. The viscid liquid would 
then be applied and the pellet of earth fitted with extreme nicety» 
Satisfied with the result of my experiment I retired. By day- 
break I found that the excavation was continued after the com- 
~ pletion of the trap door, the soil being deposited around the nest 
to raise the surface of the earth in the jar to a level with the top 
of the nest. Without close searching it was impossible to detect 
_ the door. 
= — The mode of making the trap door by this spider differs very 
_ widely from that observed by other naturalists so far as Iian 
als, construct a trap door in captivity. He made a cylindri 
= hole in a flower-pot of earth. Into this the spider disappe f : 
x During the night following the day of her capture she 
