108 THE /HUMBLE-BEE vil 
domiciles three years previously, but finer material 
brought there by mice, which had probably fre- 
quented the nests during the long interval, otherwise 
the tunnels could hardly have remained open so 
long. 
The following winter, my interest in artificial 
domiciles having revived, I determined to give 
them a much more extensive and thorough trial 
than previously. 
One of the chief difficulties in preparing a large 
number of domiciles is to get a sufficient supply of 
suitable material for the nests. My experience 
seemed to show that moss is not so suitable as fine, 
half-rotted grass, so I had some of the grass in the 
apiary mown in January, when much of it was dead, 
and in March when it was dry I stored it in empty 
hives, carefully picking out the stalks and heads, and 
cutting it with a pair of large scissors into lengths of 
about two inches. In this way I got an abundant 
supply of material, but it was far from being as fine 
as that of a mouse’s nest, and I feel sure the results 
would have been better had the material been more 
suitable. 
I placed about a dozen nests on the surface of 
the ground under bee-hive roofs, inverted boxes, 
and even tins. These nests kept fairly dry, but, 
partly I think because it is not the regular habit of 
any species of humble-bee to occupy nests in such 
positions, and partly because the material was so 
coarse, only one was occupied. ‘This nest had been 
remade by a mouse, and therefore consisted of the 
