DEDIFFERENTIATION IN ECHINUS LARVAE. 225 



of the arms shrinks, leaving the spines exposed "exactly as in 

 unhealthy larvae at all stages of development." The oral lobe 

 and the outer part of the oesophagus disappear. The inner part 

 of the oesophagus persists for a time as a completely closed tube. 

 Very remarkable changes occur in the stomach. In the late larva 

 it is highly turgid, with cells intermediate between cubical and 

 flattened. It now loses its turgidity ; the walls become very thick, 

 and eventually folded, the lumen often almost disappearing. The 

 cells are stated to multiply with great rapidity, but definite proof 

 of this is not given, and quite possibly the appearance of increased 

 number is due to the contraction of the wall ;, this point deserves 

 re-investigation. The cells round themselves off, and many mi- 

 grate into the surrounding jelly. The new stomach is reconsti- 

 tuted from the residue. 



While this has been occurring, all resemblance to an Echino- 

 pluteus form has disappeared, and the animal becomes almost 

 hemispherical. 



The essential points to be noticed are as follows : 



1 i ) The resorption of the arms. This appears to be identical 



with what we have seen in larvae placed in toxic solutions. 



(2) The dedifferentiation of the specialized larval organs, the 



epaulettes. These are not present in the earlier larva? 

 used by me. 



(3) The loss of the general form of the larval part of the organ- 



ism, and its approximation to the segment of a spheroid. 



(4) The contraction of the cesophageal tissue. 



(5) The closure of the mouth and formation of a closed vesicle 



from the remains of the oesophagus. 



(6) The contraction (loss of turgidity) of the stomach. 



(7) The migration of some of the cells of the stomach out of the 



tissues. This was paralleled in my experiments, though 

 there the cells migrated inwards to the lumen, instead of 

 outwards to the body-cavity. 



As far as the destruction of larval organs goes, we can assert 

 that the dedifferentiation caused by toxic solutions and the reduc- 

 tion at metamorphosis are closely and essentially similar. The 

 difference between the end-results is presumably due to the fact 



