228 J- S. HUXLEY. 



(2) The same is true of the related form Perophora, in spite 

 of its different mode of dedifferentiation. 



(3) In Perophora, the stolon-ectoderm is usually flat. When 

 treated with toxic agents of a certain strength, it becomes cu- 

 boidal. 



(4) In sponges, the collar-cells exposed to very mildly toxic 

 agents retract their collars. Further toxicity causes the assump- 

 tion of the spheroidal form by the cell. (Huxley, '21 a.) 



(5) In dissociation experiments, the shock of mechanical sepa- 

 ration causes the assumption of the spheroidal form, and the loss 

 of any differentiated structures such as collars, flagella, or pseudo- 

 podia, in all types of cell in sponges. (This is true also in other 

 sponges (H. V. Wilson, '07) and in Ccelenterates (H. V. Wilson, 

 'n ; De Morgan and Drew, '14).) 



(6) In the plutei here described, the gut is at first extremely 

 resistant to toxic agents. Later, however, it becomes very sus- 

 ceptible to them. This alteration in susceptibility is accompanied 

 by an alteration in appearance, the gut passing from a thick-walled 

 organ to an extremely turgid, thin-walled one. This turgidity 

 has also been noted by MacBride ('02) ; the cells in this latter 

 condition are very much flattened, with a relatively enormous 

 surface. One of the first effects of HgCU solutions in later larvae 

 is to cause a shrinkage of the gut and a separation and rounding- 

 up of some of its cells. 



It is of course presumable that high energy-consumption is 

 necessary to maintain a cell in a flattened condition, or in any 

 other involving a large surface area as would indeed be expected 

 from the laws of surface-tension. But to say that a high " meta- 

 bolic rate " thus produced is to be reckoned in the same category 

 as the high metabolic rate of a " dominant " region as defined by 

 Child, is to reduce the value of the whole very important concep- 

 tion of physiological dominance. 



Further, in cells grown in vitro, Holmes ('14) notes as a gen- 

 eral rule that any unfavorable condition leads to the abandonment 

 of an extended for a spheroidal shape. Numerous other instances 

 could be cited, but the phenomenon is so widespread as to be 

 familiar to all. 



We may therefore say that, apart altogether from the question 



