452 GEORGES BOHN 



then fasten on the shell, then the column bends in such a way 

 that the pedal disk approaches the shell. After this kind of 

 somersault, the actinian fastens itself on the shell occupied by 

 the crab. Six or seven may attach themselves successively upon 

 the same shell. 



Cornetz *'• *' has observed the following species of ants in 

 Algeria: Messor barbariis and its variety sancta, Aphaenogaster 

 testaceopilosa, Myrmecocystus cataglyphis bicolor, Tapinoma erra- 

 ticum nigerrimum, Tetramorium caespitum punicum, Pheidole palli- 

 diila. He collected with great care about two hundred paths 

 made by ants travelling alone. The return to the nest seems to 

 be a function of the outward journey; if an ant is captured at 

 the mouth of the nest and carefully transported several meters 

 away, she is unable to return, while the home journey is rapid and 

 easy when the explorer has herself made the outward trip. 

 However, the two paths, outward and homeward, are never 

 superposed; the sequence of movements and attitudes is quite 

 different on the return journey; an ant who crosses her trace 

 left on the outward trip does not stop and does not follow it. 

 The new and important fact brought to light by C. is the faculty 

 possessed by the exploring ant of maintaining her course in a 

 given direction. The insect adopts her direction on leaving the 

 nest and keeps it as well on the return journey as on the out- 

 ward trip, always resuming it after the path has been interrupted 

 either by her own researches or by the intervention of the ob- 

 server. The ant suddenly stops, makes a turn, a loop, a double 

 loop, a complicated search in the grass; or perhaps, threatened, 

 she flees in one direction or another, but soon after, she replaces 

 the axis of her body in the direction adopted on leaving the nest, 

 behaving somewhat like a compass. The error in degrees which 

 she makes in thus replacing herself is usually very slight. Ten 

 or twelve such intervals of searching about may intervene, but 

 the course remains governed by the direction taken at starting. 

 On the outward course the road is interrupted, but the same 

 orientation constantly reappears; on the return, it is still pre- 

 served. Thus the ant is brought back to a point not far from 

 her nest ; to find it, she makes the windings described by Turner. 

 Sometimes the phenomena are a little more complicated; the 

 insect outward bound adopts successively two directions, often 



