© Normally, for a swarm to depart there has to be flying bees, a flying queen and a developing queen to be. The various techniques of swarm control work (though none of them are 100% successful) by removing one or more of these factors from the equation. There are many, many different ways of achieving this, some simple, some not. It can also be dependent on what stage the colony is at in the swarming process, when it is inspected and whether the queen can be found or not.
The following method is the most commonly described but experiment with others to find one that works for you.
When the first primed cell is found carefully examine all the other brood combs for more queen cells to establish what stage in the swarming process has been reached. Queen marking pens are useful to mark the bars of those combs that have cells which saves time looking for them twice. Examine each comb, looking first for the queen herself, then for the presence of eggs. The position of the egg in the cell will indicate when the queen was last laying. Eggs still vertical in the cell have just been laid whilst those lying flat on the bottom are 2-3 days old. The queen is also often found on the comb containing newly laid eggs so when they are found have another look around for her. If no sealed queen cells are found, it is likely that the old queen is still in the hive even if she cannot be seen and no eggs are present. Unless it is obvious that about a third of your colony (the flying bees) have gone in a swarm since the hive was last inspected, the presumption must be made that the queen is still present.
So if the queen is found, put her in a queen cage and keep her somewhere safe.
This is how the hive might look before we start with the entrance at the front although there may probably be more than one brood box being used and more than one shallow. To keep the example simple the presumption is made that the deep box only contains brood and the shallow only contains either honey or a feeder. This is referred to as the old hive. The first step is to move the whole hive about three feet to the side of the original site and turn the entrance by at least 45º. The reason will become clear.
Transfer one comb containing mainly or at least some newly hatched young larvae and one comb containing mainly stores from the old hive to the new empty one. It is essential that these combs do not contain any developing queen cells; it would defeat the whole purpose of the operation so double check. The purpose of the comb containing young unsealed brood is to anchor the swarm because there is nothing to stop them from absconding but they will rarely abandon their brood. Place the comb of stores next to the wall of the hive and the brood comb next inside. If there are any empty combs in the old hive, transfer these across too where they can be used by the queen for egg laying. Put in more new frames fitted with foundation to fill the new hive. Release the queen on or near the frame of brood. With luck the queen will be found on a suitable comb which can be transferred without the need to catch her. As long as it is flying weather the flying bees will fly from the old hive but they will return, homing to this new hive over the course of the day. The new hive will then have all the flying bees and the queen but no queens to be, just like a natural swarm.Ensure the old hive has enough stores because it is losing all the flying foraging bees and has all the brood and new developing queens to feed. It is also less able to defend itself against robbers so reduce the entrance to an inch so they are able to easily guard it. The new hive, on the contrary, has very few larvae to feed but has all the flyers so if there is a nectar flow they will be able to make the most of it. If there is no incoming nectar, providing 50% sugar syrup will increase the chances of success.At its most simple, this is all that needs to be done and the colonies are left alone. With 100% success the new hive will not make more queens and the old queen will recommence laying eggs and the new queen in the old hive will emerge, mate and begin laying with new brood being visible in about 3 - 4 weeks. The hives can be united then or at the end of the season. If the queen cannot be found the procedure can be carried out anyway. The frame of young larvae that was transferred might by chance have the queen on it. If not, she will be in the old hive. Check this after a day when it will be less crowded and she may be more easily found and transferred into the new hive then.
The old hive is best left undisturbed for another 3 weeks at least unless emergency feeding is necessary in which case just check the outside store combs but do not disturb the brood nest area. It is not uncommon for disturbed hives to kill a new virgin queen. Until she has mated she will not produce the cocktail of pheromones that the workers find so attractive and they will treat her almost like an intruder during this period, biting and molesting the virgin, sometimes balling around her. At some time during the three weeks the virgin queen will hopefully emerge, take her mating flights and commence to lay eggs. If there is no sign of brood or eggs 4 weeks after splitting, put in a comb of eggs and young brood taken from the new hive or from another colony and feed some 50% syrup. This is known as a test comb as it provides a perfect test for queenlessness. The presence of young brood will often stimulate a new queen to commence laying eggs if one is present but if the colony is queenless, emergency queen cells will be found on this comb when it is next inspected in 7 days. One of these cells can be left as before or a replacement queen can be introduced or the colony can be united with another.
So, with 100% success we now have an old laying queen in one colony and a new laying queen in the other. It must be remembered that the point of this exercise is to prevent the loss of a swarm and replace an old laying queen with a new one, it is not to double the number of colonies and it can be a mistake to do so. At some point the colony should be brought back to one unit. One whole colony will collect more nectar than two half colonies and a new queen will lay more eggs in a whole colony than in a small one. When sealed worker brood, indicating the presence of a viable new queen, is found in the old hive, the old queen in the new hive is found and removed. It can be a good idea to keep her because a spare laying queen, even an old one, is a useful thing to have available. She can be housed in a retirement nucleus made from just one comb of brood and two combs of stores. It is unlikely that such a small colony will continue to make queen cells or swarm but it will still need to be checked occasionally. The old hive is then united on top of the new hive, bringing the colony back to one unit, but now with a young queen fit to head the colony for another year or two. If the old queen cannot be found, unite the colonies anyway, the workers will normally always favour a new young queen producing the maximum quantities of queen substance and the old queen will eventually be deposed.
This maintains the unity of strong colonies where large populations are more able to fight disease and gather food whilst keeping them headed by young queens that are more able to maintain the strong population with the least likelihood of swarming.
With 100% success, if the old hive was split into two, there will also now be a spare nucleus with a young laying queen that can be built up into a second colony.