XVI THE BEMBEX 221 



sandy surface. With her front tarsi, which, armed 

 with stiff rows of hairs, suggest at once broom, brush, 

 and rake, she begins to dig a subterranean dwelling, 

 standing on her four hind feet, the two last slightly 

 apart, while the front ones alternately scratch and 

 sweep the loose sand. The precision and rapidity 

 of the action could not be greater were the circular 

 movement of the tarsi worked by a spring. The 

 sand, shot backward under the creature, clears the 

 arch of its hind legs, trickling like a liquid in a 

 continuous thread, describing a parabola and falling 

 some eight inches away. This dusty jet, constantly 

 fed for five or ten minutes, is enough to show with 

 what dizzy rapidity the tools are used. I could 

 quote no second example of equal swiftness, which 

 yet in no way detracts from the elegance and free 

 movements of the insect as it advances and retires, 

 now on one side, now on another, without allowing 

 the parabola of sand to stop. 



The soil hollowed is of the lightest kind. As the 

 Hymenopteron excavates, the sand near falls and fills 

 the cavity. In the landslip are mingled little bits of 

 wood, decayed leaf-stalks, and grains of gravel larger 

 than the rest. The Bembex picks these up in her 

 mandibles, and, moving backward, carries them to a 

 distance, returning to sweep again, but always lightly, 

 without attempting to penetrate into the earth. 

 What is the object in this surface labour ? It would 

 be impossible to learn from a first glance, but after 

 spending many days with my dear Hymenoptera, 

 and grouping together the scattered results of my 

 observations, I think I divine the motive of these 

 proceedings. 



