JULY. 137 



of notice is the mode of search. This, of course, 

 varies according to locality. In digging it must be 

 borne in mind that all pupae are close to the trunk of 

 the tree, seldom more than two inches distant. Fre- 

 quently the trunk of the tree forms one side of the 

 cocoon, especially the cocoon of such insects as spin. 

 Again, pupae lie close to the surface of the earth (I 

 have seen those of Fmerinthus Populi and S. Tilice 

 half out the earth), the larva of Calocampa exoleta is 

 the only one I know of that goes to a considerable 

 depth. Insert then the trowel about three inches from 

 the trunk to the depth of two inches or so, then push 

 it to the tree and turn it up. If the soil be dry and 

 friable without grass, knock it gently with the trowel, 

 which will be sufficient. If, however, there be grass 

 you must proceed more cautiously : take up the sod in 

 the left hand, knock it very gently with the trowel, 

 and those pupae which merely enter the ground will 

 drop out. To find those which spin you must care- 

 fully examine the sod, tearing the roots of the grass 

 asunder. These are, of course, much the most diffi- 

 cult to find, the cocoons being generally the colour of 

 the earth. I may here remark. that it is useless to try 

 sticky or clayey ground, the caterpillars being unable 

 to penetrate it. In searching under moss the best 

 plan is to loosen the edge, then to tear it gently off, 

 observing whether any pupae fall. Look at the trunk 

 to see if anything adheres to it, and then carefully 

 examine the moss itself. Experience alone will enable 

 you to detect a spun cocoon. 



" If I were asked which are the most likely trees, I 

 should say those whose branches spread out towards 



