JAX B. DEMBOWSKI. 



conditions, even under the water there remains always a waste 

 chamber where air is retained and in which the crab remains 

 until the tide goes out. This process may become modified in 

 several ways. Very often Uca adjusts the sandcork in the 

 middle of the canal without rising to the mouth of it. The end 

 result remains the same, as the cork is also very tight. If we 

 pour the water a little quicker the crab works in a great hurry. 

 It rises to the top and it simply pulls many pellets of sand down 

 without adjusting them. Sometimes the animal has no time to 

 turn its large claw inward and it works hastily with II.-V. legs 

 as well as it goes. Finally, if we pour the water very quickly, 

 the burrow remains open, water enters into it, and the walls of 

 the canal collapse burrying the animal. In such a case Uca 

 usually very soon digs itself out and remains on the surface of 

 the sand until we suck the water out with a pipette. In the 

 field such an accident may be caused by a rain shower. 



These observations pour some light on the biological signifi- 

 cance of the burrow being bent and the end of it approaching 

 the horizontal line. In such a burrow the end-chamber may be 

 easier preserved from being filled with water. The air-chamber 

 sometimes may be vertical as in the Fig. i, c. But the foregoing 

 part of the burrow is then strongly bent and the end result 

 remains the same. Such air-chambers are known for several 

 sand crabs, as for Dotilla (Symons, 19) and others. 



Exactly speaking, the detecting of the biological significance 

 of the shape of a burrow does not tell us anything about the 

 factors that cause it. Beyond any doubt the direction of 

 burrowing is closely connected with gravity. This follows from 

 some experiments. In order to be able to watch closely the 

 process of burrowing I constructed a simple apparatus. Two 

 glass plates fitting closely into a jar in vertical position and 

 reaching about two-thirds of its height were put into the vessel. 

 Four corks of a suitable size (about 2 cm. thick) were put at 

 four corners between the plates and the whole was held together 

 with four threads. Thus the plates formed a kind of box and 

 the 2 cm. wide space between them was filled with wet sand. 

 Over this box standing vertically in the jar I put a round piece 

 of cardboard fitting exactly to the jar and having a cleft in the 



