David Spiro:
Actually, if you go back to the history of NYC, there was a time
in the 1800's when a "pneumatic tube" system of trains was developed,
though it did not last long.

No, there wasn't. Such a system was *proposed* and promoted by
Alfred Beach. In 1870 he opened a demonstration line about 100 m
long, with one car that shuttled back and forth into a dead-end
tunnel from a single station. Note incidentally that this was the
same year that the cable-hauled Tower Subway opened in London, also
operating a single shuttle-service car, but in this case carrying
passengers between two stations.

Both lines were equally unsuccessful. The Tower Subway shut down
(as a rail operation) in a few months. Beach kept his demonstration
line open a bit longer, but never came close to attracting the
support he'd need to build any more. At one point he gave on up
the politicians he'd originally been associated with and switched to
a different lot, and in an attempt to make this credible, put out the
story that in building the demonstration line he'd had to do it in
secret because he had *no* political backing. This falsehood found
its way into an article of mine, which I will cite in a moment.

Pneumatic tubes of various sizes *were* built in both cities for
non-passenger purposes, i.e. mail and telegrams, and I suppose
David may have had that in mind when he referred to a "system".
For more on this see my article

http://www.davros.org/rail/atmospheric.html

But for the true story about Beach, see the book-length web site

http://www.columbia.edu/~brennan/beach/

In the pictures that I have seen of it, they seemed to be about
the same width as the London trains, perhaps a bit smaller.

Quite a bit smaller. I don't think any photos exist of the interior
of Beach's single car, and drawings may not represent the size
accurately. But here are some comparative tunnel diameters:

6'8" Tower Subway
8' Beach's line
10'2" C&SLR (first deep London tube) as originally built
10'8¼" Standard early London tubes (smaller ones were enlarged)
17' Main line single-track tunnel (approximate)
25' Channel Tunnel single-track tunnel
--
Mark Brader | "...Backwards Compatibility, which, if you've made as
| many mistakes as Intel and Microsoft have in the past,
Toronto | can be very Backwards indeed." -- Steve Summit

My text in this article is in the public domain.