DEDIFFERENTIATION IN ECHINUS LARVAE. 227 



If these considerations prove to hold good, it will follow that 

 phagocytosis is a secondary phenomenon in the metamorphosis of 

 Echinoids, and probably of other groups. It will, however, obvi- 

 ously be our next task to test by experiment the hypothesis that 

 the dedifferentiation of larval tissues is the essential factor initiat- 

 ing Echinoid metamorphosis. 



C. Axial Gradients and Surface-effects. 



Child's theory of axial gradients and the differential susceptibil- 

 ity along them has been set out at length in his books ('15, etc.) 

 and papers, and it is unnecessary to enter into it here. 



I would like to point out, however, some facts which may lead 

 to a modification of some minor parts of the theory. 



Child measures and delimits his " metabolic gradients '' by 

 means of the differential susceptibility of different organs to toxic 

 agencies. In certain cases this gives concordant and uniform re- 

 sults. In other cases the susceptibility of an organ is found to 

 vary during development in a way which is not to be explained 

 without demanding a very considerable elasticity from the theory. 



The general bases of the theory appear to be founded on solid 

 enough foundations, the main difficulty being that they are perhaps 

 too general, the term " metabolic rate," for instance, being only 

 capable of application in an unanalysed sense, as an expression 

 of general total activity. But numerous exceptions, such as the 

 unpredictable variation in susceptibility of organs above alluded 

 to, can I think be explained by reference to another and simpler 

 notion. That is, that cells are more susceptible and more prone 

 to dedifferentiation in proportion to the amount of surface which 

 they are exposing. We can put this in another way, and say that 

 a cell maintains its form with greater difficulty when its surface 

 is large than when it is small. 



This statement is based on numerous facts, including the fol- 

 lowing observed by myself, as well as being deducible from theo- 

 retical considerations. 



(i) In the Ascidian Clavellwa, the parts first showing dedif- 

 ferentiation, and finally most dedifferentiated, are those where the 

 normal cells expose a great deal of surface (pharynx and atrium). 



