A cistron is a gene - here's how the word came about:
It's 1955. The proposed double helical structure of DNA has been published and the race is on to understand the implications of this for genetics. Geneticists have been working with genes for decades, but they don't know what genes are. Sanger has sequenced the two polypeptides of insulin, so the idea of a specific sequence of amino acids is established. The emerging molecular geneticists believe that genes will be explained by a feature of DNA, probably a sequence of nucleotides, but they don't know how. Crick and others are theorising about coding (i.e DNA > protein), but at this stage there is no genetic code, no mRNA or tRNA.
Seymour Benzer, at Purdue, is working on fine structure of a genetic region in bacteriophage and in this year he publishes a key paper with that phrase as the title. Without going into a lot of detail about his experimental set up, he is studying a region of the bacteriophage T4 genome called rII. Wild-type T4 can form plaques on E. coli B or E. coli K12. Mutations in the rII region cause a large plaque phenotype in E. coli B so it is easy to isolate them, but these phage cannot produce plaques on E. coli K12 which sets the scene for a powerful assay for rare recombination events.
Benzer used this system to do very fine mapping of thousands of rII mutations because he was able to use the system to measure recombination frequencies, but what matters in the present context is that he was able to do a kind of complementation test (he called this pseudo-allelism). He was able to assign the rII mutants into two groups according to whether, in a mixed infection, they produced lots of progeny. So if he mixed two phage with different mutations and didn't get many progeny he put these in the same group. On the other hand if he did get a lot of progeny he placed the two in different groups. He called this approach a cis-trans test.
"The application of the phenotype test to pairs of rII mutants leads to the division of the region into two functionally separable segments."
This is as close as he dared to go to claiming to be studying genes. He refers to 'functional units' but never uses the word 'genes' because at this stage that word belonged to classical genetics in higher organisms. In 1957 he proposed the word cistron for the functional unit defined by this type of test, and the word caught on, particularly because of its incorporation into the operon model with the idea of polycistronic mRNA. His other coinings, muton and recon were not adopted.
The two T4 genes/cistrons in the rII region of the phage genome are now called rIIA and rIIB.
Benzer, S (1955) Fine structure of a genetic region in bacteriophage Proc. Natl. Acad. Sci. USA 41:344-354