MR. ROMANES' EXPERIMENTS. 269 



which were made by Sir John that it novv' occurs to me 

 to publish them. 



"The house where 1 conducted the observations is 

 situated several hundred yards from the coast, with 

 flower-gardens on each side, and lawns between the 

 house and the sea. Therefore bees starting from the 

 house would find their honey on either side of it, while 

 the lawns in front would be rarely or never visited — 

 being themselves barren of honey, and leading only to 

 the sea. Such being the geographical conditions, I 

 placed a hive of bees in one of the front rooms on the 

 basement of the house. When the bees became 

 thoroughly well acquainted with their new quarters by 

 flying in and out of the open window for a fortnight, I 

 began the experiments. The modus oioevancli consisted 

 in closing the window after dark when all the bees were 

 in their hive, and also slipping a glass sliutter in front 

 of the hive door, so that all the bees were doubly im- 

 prisoned. Next morning I slightly raised the glass 

 shutter, thus enabling any desired number of bees to 

 escape. When the desired number had escaped, the 

 glass shutter was again closed, and all the liberated 

 bees were caught as they buzzed about the inside of the 

 shut window. These bees were then counted into a box, 

 the window of the room opened, and a card well smeared 

 over with birdlime placed ujDon the threshold of the 

 beehive, or just in front of the closed glass shutter. 

 The object of all these arrangements was to obviate the 

 necessity of marking the bees, and so to enable me not 

 merely to experiment with ease upon any number of 

 individuals that I might desire, but also to feel confident 

 that no one individual could return to the hive un- 

 noticed. For whenever a bee returned it was certain 



