94 ANIMAL INDIVIDUALITY [OH. 



seem merged and lost in the one dermal layer they 

 form, yet show themselves still independent in per- 

 forming their further function, the secretion of the 

 calcareous spicules. As these are required, single 

 dermal cells break loose from association with their 

 fellows, wander off into the gelatinous ground- 

 substance, and there take up position where the new 

 skeleton is required. Thus, though what they do 

 only has meaning in regard to the whole, the way 

 they do it proclaims them as partially independent 

 beings. 



Experiment reveals further lengths of indepen- 

 dence shows the cells capable of veritable insub- 

 ordination. By means of experiment it has been 

 possible to study the behaviour of the unit parts 

 after the individuality of the whole has been totally 

 destroyed. By chopping the sponge up small, wrap- 

 ping the bits in the finest silk gauze, and squeezing 

 them, the cells are wrenched from their attachments, 

 and pass through the meshes either singly or at most 

 by twos and threes 1 . By varying the method, one 

 can procure, instead of a mixture of all the sorts of 

 cells, a quantity of collar-cells free from all the rest, 

 and it is their behaviour that concerns us now. 



No properly conducted cell, one would have thought, 

 could wish to survive this forcible severance from the 



1 This has been done on various sponges, including Sycon, a not 

 very distant relation of Clathrina: see Huxley (9). 



