58 LOCOMOTION OF POLYCLADS 



sides of the both^ and the similarity between this mode of pro- 

 gression and the use of legs cannot escape the observer." This 

 is a type of locomotion common to gastropods and is called 

 retrograde detaxic (Olmsted, '17). This type of locomotion was 

 observed by Crozier to occur in his Bermudian species. The 

 methods of locomotion described b}^ these authors are seen in 

 the species of polyclads from Monterey Bay, but they are in 

 realit}^ four distinct methods of progression. 



In these species the swimming movement involves the whole 

 body, not merely the lateral margins as in Leptoplana tremel- 

 laria (Child, '04). P. calif ornica resorts to this method more 

 frequently than the other species. E. cavicola was never seen 

 to release itself from the substrate, and therefore never swam 

 freely. While creeping about a dish, P. californica will suddenly 

 release its anterior end, raise it above the substrate, and initiate a 

 series of waves which, from their resemblance to gastropod loco- 

 motion, maybe termed retrograde monotaxic (Olmsted, ' 17) . The 

 resulting movement is like that of a rug being rapidly shaken. 

 The waves appear at the rate of tw^o a second, and pass posteriorly 

 at such a rate that two or two and a half waves are present at 

 a given moment. An individual seldom swam more than 2 cm. 

 in this fashion before resuining its creeping. The other tw^o 

 species seldom employed this method of locomotion except when 

 falling through the water after creeping along the under side of 

 the surface film, or when somewhat roughly dislodged from the 

 substrate during active creeping. Swimming was never initiated 

 from rest, but only occurred if the animal were already in motion. 

 In this type of locomotion either the dorsal or ventral surface 

 might be uppermost, the dorsal more often in P. littoricola. 



When transverse cuts were made at any level posterior to the 

 cephalic ganglia of these pol3a"lads, this swimming movement 

 could no longer be elicited from brainless posterior pieces, but 

 the anterior pieces containing the brain, no matter how short, 

 exhibited a few swimming movements when falling from the 

 surface. Likewise after splitting the head longitudinally so that 

 the two halves of the brain were separated, an animal could not 

 be made to swim, even if one cephalic ganglion remained in each 



