BEHAVIOUR 181 



found that after 20-100 tests the animal crawled only into the 

 acceptable side. Habituation became so pronounced that Yerkes 

 stated that after these trials there was an increased readiness to 

 enter the apparatus and to desert it for the artificial burrow, an 

 apparent "recognition" of the burrow with an increasing avoidance 

 of sandpaper and stimulant, and less likeHhood of a retracing of the 

 original path along the stem of the T. The loss of the first five 

 segments, including the cerebral ganghon, does not destroy these 

 responses in a well-trained animal. As the regeneration of the head 

 end proceeds, however, the reactions of the animal become more 

 variable and less stereotyped. This continues until 2 months after 

 regeneration is complete the acquired behaviour has disappeared, 

 though the animal can be retrained again in about two weeks. The 

 majority of these results were obtained with a single, rather 

 exceptional worm, but Heck (1920) using E. foetida, L. castaneus 

 and three species of Allolobophora obtained similar data. He also 

 found that if the worms learnt one circuit which was then reversed 

 it took significantly fewer trials for them to relearn the circuit. 

 This effectively scotches the criticism that the turn to one side is a 

 stereotyped response due to a few muscle blocks or a position habit. 

 Evidently a certain amount of co-ordination in the central nervous 

 system is required. Heck (1920) also showed that not only is the 

 supra-oesophageal ganglion not essential for the maintenance of 

 the response, but the animal can also learn as readily to navigate a 

 maze without this organ as with it. It is possible to train worms to 

 turn to the illuminated side in preference to the dark by con- 

 tinually shocking them in the dark arm of the T (Wherry and 

 Sanders, 1941). The ability of earthworms to learn a maze may not 

 be of great significance in the life of the animal for the number of 

 times it comes up against such a choice repeatedly must be very 

 few. Malek (1927) has demonstrated, however, that this system 

 may be of some importance in another respect. L. terrestris (L. 

 herculeus) will drag leaves to its burrow, but if the leaf turns out to 

 be too large or stiff to get into the hole the attempt to draw it down 

 ceases after ten to twelve trials. The observation that certain types 

 of leaf are eaten in preference to others, and that physical shape 

 and texture may be important in governing the choice has also been 

 noted by Wittich (1953). 



Yerkes (1912 a, b) remarked that there is a daily fluctuation in 



