In the book, "The Intention Experiment," a professor hooked up a plant to a polygraph machine. He then had the intention (in the form of a thought) to burn one of the plants leaves.
Cleve Backster was doing this in the 60s. His experiments with yoghurt communications are fascinating, although kinda out-there for most!
The thing i find most intriguiging is that maybe it was the very intense thought it self could have triggered the polograph to give the odd results.
But this thoery arises the question as of "why criminals have a hard time to passing poloygraphs..." well it could be they didnt have enough conciousness(thought power) in the first place to find alternative ways to accomplishing the negative thing they are getting accused with.
I read Backster's book a few years ago. There are some interesting interviews of him online, such as this one:
http://www.derrickjensen.org/backster.html.
Polygraphs are notoriously unreliable. Lots of interesting reading at
http://antipolygraph.org/. The success rate is slightly above chance, and that seems to be mainly due to the examiner scaring the subject into confessing! Do you know the (lying) basis on which a polygraph test is done? There's a
220-page PDF file at that site which explains all.
Like similar biofeedback devices, they show momentary stress in the subject, such as a reaction to a hot question. A real remorseless criminal will not have a stress reaction to what you think might be a hot question, like "Did you steal the unsecured wallet.dat file?" He may have a reaction to a question that violates *his* moral code, like "Did you fail to steal the wallet.dat files that you could have done?" if he feels bad about missing out on the opportunity. But polygraphs are not done on this basis, but on the basis as described in that PDF. Interesting subject.