234 FORTY YEARS AMONG THE BEES. 



from the time the e^g is laid, and is fed throughout her 

 larval lifetime on the same food that is given to a 

 worker-larva during the first three days of its larval ex- 

 istence. So a worker-larva more than three days old, or 

 more than six days from the laying of the egg, would be 

 too old for a good queen. If, now, the bees should 

 select a larva more than three days old, the queen would 

 emerge in less than nine days. I think no one has ever 

 known this to occur. 



BEES DO NOT PREFER TOO OLD LARVAE. 



As a matter of fact bees do not use such poor judg- 

 ment as to select larvae too old when larvae sufficiently 

 young are present, as I have proven by direct experi- 

 ment and many observations. It will not do, however, 

 to conclude from this that all queen-cells started by a 

 queenless colony left to themselves will be equally good. 

 Bees have a fashion of starting cells for a number of 

 days in succession, and will continue to start them when 

 larvse sufficiently young for good queens are no longer 

 present. So some means must be taken to make sure 

 that no nucleus has for its sole dependence one of these 

 latest cells. If several cells can be afforded for each 

 nucleus, there is little danger they will all be bad. 

 Neither is there great danger if a cell is chosen which is 

 large and fine-looking. Perhaps the safer way is to 

 give the queenless colony a frame with eggs and young 

 brood three or four days after the removal of the queen, 

 and then they will not be obliged to use the older larvae 

 of the other combs. 



PLACING QUEEN-CELLS. 



Two or three frames of brood with adhering bees 

 are taken for each nucleus. If one of the frames has a 



