Here's a good piece on N-rays and several similar cases of apparent self-delusion:

In 1903, Blondlot, who was a well-thought-of French scientist, member of the Academy of Sciences, was experimenting with x-rays as almost everybody was in those days, The effect that he observed was something of this sort. I won't give the whole of it, I'll just give a few outstanding points. He found that if you have a hot wire, a platinum wire, or a Nerns't filament or anything that's heated very hot inside an iron tube and you have a window cut in it and you have a piece of aluminum about 1/8 of an inch thick on it, that some rays come out through that aluminum window. Oh, it can be as much as two or three inches thick and go through aluminum, these rays can, but not through iron. The rays that come out of this little window fall on a faintly illuminated object, so that you can just barely see it. You must sit in a dark room for a long time and he used a calcium sulfide screen which can be illuminated with light and gave out a very faint glow which could be seen in a dark room. Or he used a source of light from a lamp shining through a pinhole and maybe through another pinhole so as to get a faint light on a white surface that was just barely visible.

Now he found that if you turn this lamp on so that these rays that come out of this little aluminum slit would fall on this piece of paper that you are looking at, you could see it much better. Oh, much better, and therefore you could tell whether the rays would go through or not. He said later that a great deal of skill is needed. He said you mustn't ever look at the source. You don't look directly at it. He said that would tire your eyes. Look away from it, and he said pretty soon you'll see it, or you don't see it, depending on whether the N-rays are shining on this piece of paper. In that way, you can detect whether or not the N-rays are acting.

Well, he found that N-rays could be stored up in things. For example, you could take a brick. He found that N-rays would go through black paper and would go through aluminum. So he took some black paper and wrapped a brick up in it and put it out in the street and let the sun shine through the black paper into the brick and then he found that the brick would store N-rays and give off the N-rays even with the black paper on it. He would bring it into the laboratory and you then hold that near the piece of paper that you’re looking at, faintly illuminated, and you can see it much more accurately. Much better, if the N-rays are there, but not if it’s too far away. Then, he would have very faint strips of phosphorescent paint and would let a beam of N-rays from two slits come over and he would find exactly where this thing intensified its beam.

Well, you'd think he'd make such experiments as this. To see if with ten bricks you got a stronger effect than you did with one. No, not at all. He didn't get any stronger effect. It didn't do any good to increase the intensity of the light. You had to depend upon whether you could see it or whether you couldn't see it. And there, the N-rays were very important ...

http://www.cs.princeton.edu/~ken/Langmuir/langB.htm