THE METHOD OF TRIAL AND ERROR. 245 



by Miss Frances Dunbar, which I hope may soon be published, show 

 that that animal finds its food in a similar manner. All searching is, 

 of course, behavior on the plan of trial and error, and many organisms 

 are known to search for food. 



The righting reactions of organisms are among the most striking 

 examples of trial and error in behavior. In the starfish, for example, 

 when the animal is laid on its back " the tube feet of all the arms are 

 stretched out and are moved hither and thither, as if leeling for some- 

 thing, and soon the tips of one or more arms turn over and touch the 

 underlying surface with their ventral side" (Loeb, 1900, p. 62). As 

 soon as these one or two arms have been successful, the others cease 

 their efforts ; the attached arms then turn the body over. If all the 

 arms attempted to turn the animal at the same time, in other words, if 

 there were no way of recognizing " success " in the trial, the animal 

 could not right itself. 



The righting reaction of the starfish shows much resemblance to the 

 method, described above, by which a suspended Amoeba passes to a 

 solid. It is probable that a Diffltigia, turned with the opening of the 

 shell upward, would show a righting reaction essentially similar to 

 that of the starfish. 



The righting reaction of the flatworm Planaria, as described by Pearl 

 (1903), is not so evidently brought about through the method of trial 

 and error. Yet there are certain facts that indicate that this method is 

 really essentially present here. Thus, Pearl shows that when the flat- 

 worm is prevented from righting itself in the usual way, its rights itself 

 in another manner. Probably various reactions are tried ; if the first 

 does not succeed another may. 



This peculiar form of the method of trial and error, in which several 

 different reactions are tried even under but a single stimulus, is brought 

 out by Mast (1903) in the behavior of Planaria under other conditions. 

 If the water containing the flatworms is heated, the animals give, as the 

 temperature rises, practically all the reactions that they ever give under 

 any conditions. 



We have thus, as the temperature rises and the stimulation increases, the fol- 

 lowing reactions given consecutively: positive, negative, crawling, righting, 

 and final (all the reactions described by Pearl, with the exception of the food 

 reactions, and the final reaction in addition). (Mast, 1903, p. 185.) 



We shall have occasion to inquire as to the significance of this 

 responding to the same stimulus by many different reactions when we 

 take up, in another connection, certain similar phenomena in Stentor. 



In the higher vertebrates, as we have mentioned at the beginning, 

 the method of trial and error plays a very large part. It is here espe- 

 cially that it has been recognized as a definite type of behavior in the 



