First Attempts at Observation 



centre of the triangular base. A small 

 earthenware pie-dish with a hole made in 

 the bottom, receives the open upper end, 

 which projects a Httle and holds a layer of 

 earth that comes level with the brim. This 

 will represent, around the mouth of the 

 shaft, the space in which the insect can attend 

 to its business, either to shoot the rubbish 

 from the shaft or to gather the provisions 

 round about. Lastly, a glass bell, fitting 

 into the dish, prevents escape and preserves 

 the slight quantity of moisture needed. A 

 few supporting strings and bits of wire keep 

 the whole thing firmly fixed. 



We must not overlook one most im- 

 portant detail. The diameter of the tube is 

 about twice that of the natural burrow. 

 Therefore, if the insect digs along the axis 

 and in an exactly perpendicular direction, it 

 will have at its disposal more than the re- 

 quired width. It will obtain a channel lined 

 on every side by a wall of sand a few milli- 

 metres thick. We may however assume 

 that the digger, knowing nothing of geomet- 

 rical precision and ignorant of the condi- 

 tions provided for it, will take no account 

 of the axis and will deviate from it to one 

 side or the other. Moreover, the least ad- 



lOI 



