The Mason-bees 



capable of cutting a thin membrane? This 

 is tlie point to look into first of all. 



In February, by which time the insect is in 

 its perfect state, I take a certain number of 

 cocoons, without damaging them, from their 

 cells and insert them each in a separate stump 

 of reed, closed at one end by the natural wall 

 of the node and open at the other. These 

 pieces of reed represent the cells of the nest. 

 The cocoons are introduced with the insect's 

 head turned toward the opening. Lastly, my 

 artificial cells are closed in different ways. 

 Some receive a stopper of kneaded clay, 

 which, when dry, will correspond in thickness 

 and consistency with the mortar ceiling of the 

 natural nest. Others are plugged with a 

 cylinder of sorghum, at least a centimetre^ 

 thick; and the remainder with a disk of brown 

 paper solidly fastened by the edge. All these 

 bits of reed are placed side by side in a box, 

 standing upright, with the roof of my mak- 

 ing at the top. The insects, therefore, are 

 in the exact position which they occupied in 

 the nest. To open a passage, they must do 

 what they would have done without my inter- 



1.39 inch. — Translator's Note. 

 34 



