I'm reading A Random Walk in Science, a compilation of science humour from rather ancient times to shortly before the book was published, in 1973.
It contains some, but not many, things I've seen before - the turboencabulator, The Contributions of Edsel Murphy to the Understanding of the Behavior of Inanimate Objects, A glossary for research reports et cetera.
(legible version below)
My favourite bit so far, though, is The art of finding the right graph paper to get a straight line, from an almost-fifty-year-old volume of the Journal of Irreproducible Results.
This piece is not on the JIR Web site (though this other excellent graph is), and it doesn't seem to be online anywhere else, except for this site that lets you read a who-knows-how-legal copy of the whole book. (Or of course, you could download the book from a hive of scum and villainy.)
A Random Walk in Science is also still in print, too, though ridiculously expensive. So I've taken the liberty of image-ifying those two pages. Click for more legible versions.
This is probably still copyright to somebody, no warranty expressed or implied, et cetera.
I also rather like this quote from Sir Arthur Eddington:
When an investigator has developed a formula which gives a complete representation of the phenomena within a certain range, he may be prone to satisfaction. Would it not be wiser if he should say "Foiled again! I can find out no more about Nature along this line."