204 THE ESSEX NATURALIST.
horizontal mirror ; except when under examination the tubes
were kept in the dark.
Photographic records were made of the various changes which
were seen to occur from time to time. They show a consecutive
series, not all from one example, but chosen from a number of
photographs of all the tents. To get a similar series from a single
individual would require constant watch over the whole opera-
tions from start to finish, occupying 5 or 6 days. I have not seen
the actual start of the making of a tent, but I was fortunate in
finding one in the garden not long after it had been begun. It
was photographed at once, and then the gong sounded for lunch
and I left the millepede busy with its work, under the camera,
expecting it to carry on whilst I made records of its progress
during the afternoon. On my return, however, I found that the
millepede had eaten a small hole in the unfinished tent, and had
escaped, and I saw it no more, in spite of a careful search.
On another occasion one of the captives was seen to be busy on
a recently started tent at 7 a.m. I photographed it at once and
took the precaution this time to put it back into the tube, of course
placing it with the tent on the underside of the wood as it had
previously been hanging One hour later, however, I found
the tent deserted. A second start was made by the creature dur-
ing the day and the tent was seen to be completed when inspected
about 6 p.m. The silk appeared to be drawn out from an aper-
ture below the head, but the exact place requires further investi-
gation.
The millepede evidently objects to any interference whilst at
its work, at least during the early part of the operations.
The position and structure of the silk glands and their ducts
of course require further study.
One tent was on the underside of a small block of wood and
when a piece of this along a crack at right angles to the surface
was removed another tent was unexpectedly found in the crack,
which in this case was of course spun with its base in a plane per-
pendicular to the surface of the ground.
A tent (Plate X, fig. 2) was found on August 20. Two days
later, at about 11.45 a.m., it was seen that the inmate
was in the process of moulting (Plate X, fig. 3). The white
skin was being slowly sloughed off at the tail end. The
millepede rotated as the white skin lengthened, the latter simply