older books

>100% on that one!!!  Books,  (often the older ones from 1850-1950 give the
>info even if a bit dataed), are a prime source of good data.  The age of raw
>impericism was in full swing and authors felt obliged to give concrete
>photographs, diagrams or drawings, explanatory text, and sample problems
>associated with what they covered rather than pages of derivational,
>math followed by some convoluted, multi-leveled quiz.  The math the old
>did give was directly usable to engineer or work up approximations for real
>stuff an often came with a solid dose of caveats warning of where the
>broke down.

100% complete agreement :)
I don't even own an engineering book printed after 1940. After two years of
trying to learn engineering "code," and crashing and burning in the attempt,
I started finding older texts at flea markets and thrift stores, etc. Now I
have a collection of engineering texts that teach/taught me in plain
non-exclusionary language! My personal belief is that those older texts
were written by people that actually wanted to *teach* something, rather
than exclude others, or ego-trip by numericizing in the most obscure
mathematical "Greek" possible. In defence of electrical engineering
however, I also have a friend that didn't crash and burn in the curriculum,
and his grasp of theory has helped filling in the gaps. :) 
Not only do the books speak in plain language, but one also gets a good
grasp of the historical implications of what one reads therein. For what
it's worth, the ham-radio newsgroups are good places to shop for older
texts. Just ask for a particular book and almost always, some old-timer
will have an extra copy of some excellent text for a very good price :)
My two cents,