As you can see, the thatching was slowed a little by a 6-inch snow. 
The thatch was wrapped over the perlins,
and tied like this,
so the next course would cover all the lashings and laps.
That made the interior look like this.
On the corners, loops
were made for a place to attach the thatch once all the perlin space was used.
Long poles were stuck through the perlins to hold bundles of thatch.
With
this method, the thatch was easily grabbed by reaching around the corner then
applying it to the adjacent side. 
One of the goals of this experiment was to see if such a
small structure could hold much weight, such as a 220 pound person or a heavy
snow load.
This should answer that question. Here,
Larry stacked bundles of thatch within reach to facilitate
their installation. 
For the second-last course, this loop was made
then thatch was applied to the loop
and it was slipped over the top and fastened into place. The
peak always seems to present a problem when thatching is applied. Here it was
solved by tying a bundle of thatch near the larger end and placing it on the top
of the peak.
Thatching the entrance or "keyhole". The transition problem, where the main
structure's thatch met the "keyhole" thatch, was never really satisfactorily
resolved.
That
resulted in leaks and quickened the demise of the structure. 
If
Larry were to build such a structure again, he would cover the "keyhole with
bark and let the thatch from the main structure overlap the bark. The problem
with the thatching was that there wasn't enough room to tie the thatch in place
where the "valley" met. That, in turn, allowed the "valley" to leak.
This is what it looked like in1990. It is collapsing
downhill, to the right.
The keyhole structure stood for 2 more years and, as of 2009, is now a small depression in the ground. It is hoped that, some day, someone will re-excavate the site.
