This replies to the 26th January posting by Vladimir Obolonkin.

The original spectrum analysis was carried out 50 years ago by E.F.Woods who
detected many sounds in the hive.  Only two were useful in predicting
swarms.

1)  A warbling or bubbling sound in the band 225-285Hz, made by unemployed
nurse bees and indicating that the queen has reduced her egg production.

2) A hissing sound lasting about a second that occurs  when the hive is
bumped.  A short sharp hiss indicates that the bees are loyal to the colony
and not planning to leave.  A quieter, more 'rounded' and slightly longer
hiss indicates that they are losing their loyalty and preparing to depart.

Eddie designed an instrument which he called an apidictor which was a simple
amplifier with a microphone and earpiece.  Holding the microphone to a hole
in the back of the hive you switched through three sounds.
One was the complete hive noise.  The second blotted out everything but
225-285Hz which was allowed through a band pass filter.  The third blotted
out everything except those above 3,000Hz.  This emphasised the hiss and
made it easier to distinguish between the two types mentioned above.

The apidictor was a simple robust instrument designed for a beekeeper to
carry round on his inspections to help decide which colonies were going to
swarm.  It meant that instead of inspecting them all every 8 days, he only
had to inspect 10%.  This not only saved his time but avoided losing honey
through the disruption of inspecting the other 90%.

It was too technical for the beekeepers of the day but it was certainly not
a toy.  Made in 1964, there are many still in use.

There is nothing wrong in using your computer as a spectrum analyser for the
same purpose but, if you can understand this English metaphor, it is 'using
a sledgehammer to crack a nut'

If Vladimir would like more explanation, I hope he will e-mail me direct.

Rex Boys.