One 19th century pipe study is by Alexandra Dane and Richard Morrison
[1979], Clay pipes from Port Arthur 1830-1877.

D and M didn't measure stem diameters citing Oswald and Walker as saying the
Binford regression formula breaks down in the 19th century, with Walker
saying this is due to:

1.    mass manufacturing and proliferation of different mould types

2.    bores less than c. 4/64" impractical with available technology

3.    increased use of shorter pipes.

Reason no. 2 sounds the most intuitively convincing to me.

Denis
----- Original Message -----
From: "George Myers" .....>

......
> So there are pipestems from the nineteenth century that could play with
the
> formula, and the known variance in the Dutch pipestems that contradicts
the
> standard regression hypothesized for most pipestems. I am not aware of
19th
> century pipestem studies.
>
> George Myers
>