NERVOUS SYSTEM OP ANTEDON ROSACEUS. 525 



when otherwise specified, both needle and forceps irritation 

 were tried. In some instances the application of acid by a fine 

 brush was made use of as an irritant ; but this can only be 

 done satisfactorily on specimens removed from the water. 



Experiment 3. — An active specimen was eviscerated and 

 allowed to come to rest in the normal position. It was then 

 inverted and placed mouth downwards on the bottom of the 

 tank. After a short rest it righted itself in the normal manner, 

 but rather more slowly than usual, the interval between in- 

 version and completion of the righting manoeuvre being about 

 two and a half minutes. This experiment was repeated many 

 times with different specimens. Some righted themselves 

 instantaneously, others took a longer or shorter time, but the 

 general average of the times taken by eviscerated specimens to 

 right themselves was about half a minute longer than that of 

 uninjured ones. 



This affords strong additional evidence that the co-ordinatins: 

 centre of the complex muscular movements of which an 

 Antedon is capable is situated not in the visceral mass, but in 

 the calyx. 



C. On the Power of Regeneration of Eviscerated 

 Specimens. 



It has been stated above that an eviscerated Antedon not 

 only attaches itself by its cirri in a perfectly normal manner, 

 but that it may remain so attached for a week or more. On 

 experimenting one day with a specimen that had been evis- 

 cerated about a fortnight previously, I noticed that it righted 

 itself when inverted rather more readily than is usual in evis- 

 cerated specimens ; and on examination I found that very con- 

 siderable regeneration of the visceral mass had occurred. The 

 soft tissues lining the calyx were of some thickness ; a mouth 

 was already present in the centre of the oral surface, and 

 ambulacral grooves had formed converging from the arms 

 to the mouth. I at once took steps to secure a complete 



